Long Endurance/Range

[wowo44]

Have any of you thought about possible solutions to enable a UAV to have long range and endurance? I hadn't really thought about this until recently, but it seems like it would be quite an interesting challenge.

What are some of your ideas for maximizing the endurance of a small UAV (gas/glow or electric)?
How would you implement telemetry for a long-range UAV?

I haven't thought too much about maximizing endurance; this seems like the most challenging aspect. Oviously, if the UAV is going to be venturing hundreds of miles from home the attitude reference must be able to function in IMC. So this means no infrared horizon sensing.

For telemetry, it could be possible to use cell phone technology. This would limit telemetry to areas within range of a cell tower. It could also prove to be expensive (long-distance charges).

I can imagine a ground station that periodically connects to the UAV for telemetry updates. To remain economically feasible, it would not be real-time.

What are some of your ideas?

[minifly]

look at http://www.edgerc.com/Locust%20UAV-MAV.htm

[danstrider]

Ya know, endurance is one of those tricky things. Looking at the long-endurance uav's already out there, seems like some monster batteries (like the aerovironment mav) or a glider-like planform (like predator, global hawk, pointer, raven, etc) is needed from the get-go. Even our supposed high-endurance models don't seem to compare to the military or commercial requirements.

So beyond the platform design, telemetry is a beast. Without satellite links, seems like a cell phone link is about the only way for a modeller to go. LOS links aren't always feasable unless you're following your own car convoy or something. What else is there?

Dan

[LukeZ]

Long endurance (or long distance, not always the same thing) is definitely a challenge. NASA's Helios comes to mind, which was essentially a massive flying solar panel. It travelled very slowly, however, and worked best at extremely high altitudes. I've looked into solar and my personal conclusion just for me is that it's really not very practical for the hobbyist, at least for long distance. Aside from being expensive, solar panels are fragile, and not very efficient. There have been several solar powered r/c gliders, but none that I know of that could store enough energy during the day to fly through the night. So although solar might be a good way to augment the energy supply on an endurance aircraft, I personally don't think it's practical as the primary source, at least for someone on a limited budget.

On the other hand, solar is about the only perpetual source of energy we could expect to use, so if it's not the primary power supply then by definition any endurance aircraft is going to have to come down eventually (even a solar aircraft would as well, technically speaking, but theoretically it wouldn't have to). That's kind of a bummer, as the idea of an aircraft that could theoretically stay aloft indefinitely really is cool. (I guess a lighter-than-air ship would be the only other type I could think of. Hmm... :rolleyes: )

The Aerosonde (http://www.aerosonde.com/) UAV can fly quite a ways, about 3,000 km I believe. But that's just because it has an efficient engine and a big gas tank. If you want to go that route you can basically go as far as you want, as Steve Fosset just showed with his Global Flyer. But I don't find that very challenging personally. It's like taking a gas powered boat across the Atlantic: I think it's more exciting to sail. And here's a guy (http://news.nationalgeographic.com/news/2005/04/0407_050407_solarairplanes.html) trying to do just that- fly around the world in a solar powered aircraft.

Personally I'm interested in long distance/endurance electric powered aircraft. LiPolys are great but I'm looking forward to the next generation beyond that, such as fuel cells. Another thing that could extend the longevity of the flight would be the ability to detect and take advantage of thermals. Additionally, if the specific route from A to B wasn't necessarily important, the aircraft could be fed current meteorological data and have it decide which route would take the most advantage of prevailing winds.

As for telemetry, I think cellular is probably the best (though not ideal) option for the average person. A satellite phone could also be used but besides the expense, it seems to me they're all pretty heavy.

If you had enough money you could build and have launched your own CubeSat (http://cubesat.calpoly.edu/_new/index.html) satellite. I don't know if they do geo-synchronous orbits or not, but that would be best. I think the price is anywhere from around $40K on up. Maybe if enough of us got together and pitched in? :p

[ElectroLawndart]

If you had enough money you could build and have launched your own CubeSat (http://cubesat.calpoly.edu/_new/index.html) satellite. I don't know if they do geo-synchronous orbits or not, but that would be best. I think the price is anywhere from around $40K on up. Maybe if enough of us got together and pitched in? :p

That's WAAAAY better that dropping my glider from a high altitude balloon. Now if I could just remember where I put my recipe for silica tiles. :D

[LukeZ]

According to the rules (http://cubesat.calpoly.edu/_new/documents/cubesat_spec.pdf):

All parts must remain attached to the CubeSats during launch, ejection and operation. No additional space debris may be created.
Darn! Some party-pooper already rained on your parade... :rolleyes:

But a glider isn't really space debris, is it? :)

[AnthonyRC]

One project that I've been thinking (ok, dreaming) about for some time is an autonomous slope-soaring machine.
Assuming first that you live near some decent mountains. Give the UAV some knowledge of the terrain, a topographical map. Then give it the ability to sense wind direction and speed (fairly easy to do, even by just circling and watching GPS position).
You could launch it from a slope, climb to a reasonable altitude, and then hit the 'autopilot' switch. The plane would then keep climbing, and then pick another suitable slope which is on route to it's chosen destination, fly to it, and do the same.

With some reasonable LiPos powering the receiver/microcontroller/gps, it would appear that you could stay up as long as you have wind.

Obviously this only works where you have mountains, but that covers a fair chunk of the globe.

Any comments/thoughts?

[danstrider]

I read someone's brainstorming about this exact idea on the sourceforge autopilot group. I did not keep up with his progress, but it sounded feasable...

http://sourceforge.net/mailarchive/forum.php?forum_id=296&max_rows=25&style=flat&viewmonth=200210

Dan

[ElectroLawndart]

One project that I've been thinking (ok, dreaming) about for some time is an autonomous slope-soaring machine.

Any comments/thoughts?

Yeesh! An autonomous slope glider would be daunting indeed. I would think just the number of sensors would be prohibitive. Just off the top of my head...
GPS - obviously.
Altimeter - good idea, a 1 update/sec. might be a bit slow flying close to terrain.
Radar(Sonar?) Altimeter - need to know how high above said terrain you are.
Artificial horizon - with all the pitching and banking a slope glider does, would sensor drift be an issue? Could IR horizon sensing work when half you horizon is obscured by terrain features?
Air Speed - good
Heading - hmmm. slopers are mostly flying at some kind of crab angle.

...and most of all.
How does the UAV recognize slope lift.

I've been thinking of doing some research into hybrids by taking a .049 converted to diesel and mating that to a brushless generator and NiMH packs to run a small brushless DC motor coupled to a variable pitch prop. At takeoff the propulsion motor is run off the generator and battery packs. Once up to altitude, the diesel is shut down and the prop runs off the just the batteries. The generator just starts up to top off the batteries.

Dart

[sesat]

UAV's estimate the velocity of wind with respect to ground using sensed information about ground velocity (wrt to uav body) and airspeed (wrt to body).


Ram.

[ElectroLawndart]

sesat,

Could you elaborate?

Dart

[wowo44]

Sesat is referring to the "triangle of velocities". Basically you have two vectors:
1) airspeed and heading
2) groundspeed and track

the resultant of these two vectors is the wind speed and direction.

[LukeZ]

Dart, I have been thinking of the hybrid route myself, but have not done any testing yet. I need a brushless motor first. I've read abou a few people who have made generators from them here on RCGroups but I don't think any of them posted performance figures.

For an endurance plane I've wondered how well it would work though. You're going to have big batteries, which means the need for an even bigger generator. Let's say I could build a rather slow flying glider UAV that can travel along at say 15 amps current consumption, including the power to onboard electronics. That's a pretty conservative figure I'd say. Let's say you wanted four hours of continuous operation from the batteries before they needed a recharge- that's still six times a day you'll have to fire up the engine. 15 amps for four hours is a 60 amp Lipoly pack. Not impossible by any means when you stack them up in parallel. But now, to recharge them? The generator is going to have to put out 60 amps for an entire hour, or some amount less for longer (I'm talking crudely here). Let's say your generator could only charge them at 1/2C, then it would need to run for two hours a charge. Six charges per day, two hours per charge, you're already up to running the engine for half of the day. At that point I start to wonder if I shouldn't have just had the engine be the primary power source to begin with.

Any attempt to reduce the number of charges per day (I'm talking about an endurance aircraft that can run for several days), will involve using higher capacity battery packs, but then that in turn means they take longer to charge.

And of course this isn't even getting into the issue of how many times a Li-Po can safely be recharged consecutively like that.

In actuality you would probably want to have two battery sets, one could be powering the plane while the other could be charging or resting. Solar could also be used to augment the available supply, but it won't help much.

At any rate, what's really needed are some performance figures from these generators. If an .049 or other small/fuel efficient engine could spin a massive brushless, or an array of smaller brushless motors, and create huge amounts of potential energy for very little fuel, then it might work.

It may also be possible to wind your own brushless motors that would be optimized for energy output. I'm not the brushless guru, but my guess is that at present, manufacturers would try to maximize the performance of their motors for the minimum amount of energy input - to the extent that a brushless can even be optimized for such things. But if so, that would be the opposite of what is desired when you're using them the other way around.

Just my thoughts on that topic. It's probably worth its own thread but I'm not going to start one until I have some numbers to show: anyone else who does, though, feel free. I'd be really interested to hear what others have found.

[ElectroLawndart]

LukeZ

I was thinking more like 5Amps max to sustain flight. The variable pitch prop would be needed for efficiency. Roll the motor RPM's back and crank up the pitch. Possibly flying a powered glide flight profile. Long powered glide on the brushless, start up the generator to charge the batteries and augment power to the brushless for a climb back up to altitude. The generator would act as the starter by putting power to it.

I would be using either NiCad's or NiMH batteries because of the ability to fast charge them.

The airframe would be along the lines of a soaring glider with a big folding prop on the front. I don't know if I could get the L/D out of the standard pod and twin boom UAV configuration.

I haven't done much thinking on this. A project like this would be at least 2yrs out for me. Still trying to design and build a balloon drop glider.

Dart

[sesat]

Thanks Wowo.

One hope I harbor for long distance or long duration flying is a glider that effeciently rides thermal after thermal. A robot LSF flyer :eek:


Ram

[ElectroLawndart]

wow44

I understand that the offset between the groundspeed vector and airspeed vector is your windspeed vector but how would your vehicle find slope lift. Would you take your windspeed vector overlay it over a map of the ridgeline your flying and decide if there is going to be lift there or not. If so, then all the ridges along your flight plan would have to be mapped.

Dart

[AnthonyRC]

Some interesting ideas!.
To answer the questions about where lift will be expected, the idea was fairly simple, if the microcontroller in the plane (which would need to be fairly 'capable') had access to a detailed map, which inclues altitude information, then it should be able to predict it. Some pre-processing of the map on a PC before loading into the plane would simplify this, you probably want to avoid slopes that are covered with trees.
Maybe in areas with some decent mountains this high-altitude 'mountain wave soaring' could be used too: https://ntc.cap.af.mil/ops/DOT/school/NCPSC/GliderNCPSC/CAPF_5_glider/soaringtechniques.htm

[lvmikede]

WE had some guys working on one here at school and they modified a 6m glider. They embeded solar cells all over the plane and ran that to a converter to charge a ccd camera, and the flight batteries. I think it was launched with a Jato setup or tow.

[danstrider]

Do you have some pics lvmikede? I'd love to see them!

What school is this?

Dan

[sneu]

Here is a picture after Alan Cocconi's plane after 24 hours of constant flight last week. It was powered by 220 watts of solar cells and 10 lbs of lithium ion batteries. It was flown on El Mirage dry lake in Southern California. Plane weighed 25 lbs and has a 15 foot wing span. The entire flight was made from the small trailer in the background. Takoff power is about 800 watts and cruise power is 50-100 watts. The flight started at 12:30 am and landed the following night after 12:30 am.

Steve

[danstrider]

Wow that's so cool! 24 hours!!! Is it a custom plane (obviously mounts & stuff are custom)? Does this project have more information on a website? More closeup pictures of the solar cells? Circuits? Autopilot or video system onbord? I'm stoked, that's a pretty awesome project *BIG GRIN*

Dan

[ElectroLawndart]

Hey Sneu,

Thats great. When was the flight made? Just like DanStrider said, any information would be appreciated.

Dart

[copes]

Electric Guru Alan Cocconi's site:

http://www.acpropulsion.com/

[sneu]

Wow that's so cool! 24 hours!!! Is it a custom plane (obviously mounts & stuff are custom)? Does this project have more information on a website? More closeup pictures of the solar cells? Circuits? Autopilot or video system onbord? I'm stoked, that's a pretty awesome project *BIG GRIN*

Dan


Alan made the entire plane and most all the electronics for it also. He even designed a new brushless controller that far out performs the usual ones that are used for most model applications. While most controllers are pretty good at full or near full power the system efficiency falls off very fast as the power levels drop. In order to have a chance at the 24 hour endurance he needed something better and he set out to build it and did. the plane used a Kontronik Tango motor and a 9 phase controller. At full power the system efficiency is over 90%--at 10% power it declines less than 5%. Operating at the same power with "normal" controllers would yield a reduction of efficiency on the order of 40% or more.

The solar cells were molded into the upper surface of the wing. I have some more pictures that I can post later.

I did a few hours of flying for Alan during the early afternoon of the 21st of April. It was a treat to fly thermals from video with a full set of flight instruments and GPS.

Steve

[LukeZ]

Steve,

I think it's just great that you've posted here. Obviously Mr. Cocconi has already reached the place that a lot of us are presently shooting for.

Any more pictures and technical details you can provide I assure you will be devoured!

[radiohound]

That was a great article (located here) http://www.acpropulsion.com/ACP_PDFs/AC_Propulsion_Solong_UAV.pdf

which shows some of the planes specifications. It got me very interested in the solar cells he used which look very thin and light. They are also 20% efficient which brings them up to par with what Nasa is using in space. Most consumer available cells get about 12 to 15% efficiency.

I called to get a price, just to see if the prices were astronomical, but it turns out they are $12 a piece if you buy the minimum 50 cells. That's actually a lot more reasonable than I thought they would be. This plane used 76 cells.

Here is their spec sheet, just in case anyone else was curios.

http://www.sunpowercorp.com/html/Products/datasheet/A-300/A-300.pdf

[danstrider]

hey, that's not too bad.

The pdf was pretty doggone cool, an excellent read :-)

I think my next project will include solar something...

[LukeZ]

Well, that's about a grand for the solar cells, and pretty much a grand for the batteries (100 LG Chem 18650 Li-Ion). The price on the solar cells are more reasonable than I had anticipated too, but I'd say that only barely brings them into the margins of what a serious UAV hobbyist could afford. Still, the ability to fly essentially indefinitely?! Almost worth it!

Molding the solar cells into the surface of the wing like he did is really slick. I wonder if that's a very difficult process. He did make an alumimum (!) wing mold, that might answer my question right there... :rolleyes:

[gnz]

which shows some of the planes specifications. It got me very interested in the solar cells he used which look very thin and light. They are also 20% efficient which brings them up to par with what Nasa is using in space. Most consumer available cells get about 12 to 15% efficiency.
http://www.sunpowercorp.com/html/Products/datasheet/A-300/A-300.pdf

RWE Space Solar Power Gmbh has also some efficient GaAs solar cells (average efficiency 26.6%)

http://www.rwespace.de/pages/products_gaas.htm#

[lvspark]

might do some checking on Arotech zinc-air batteries.

Also a link to solar thinfilm http://www.iowathinfilm.com/products/index.htm

[RCAV8R13]

I built and piloted the take off of the Aerosonde "Laima" that flew across the Atlantic in 1998. She landed with 20% fuel reserve and flew 2,000 miles. She got about 1,350 mlies to the gallon! It took 26hrs 45min.
Our new UAV, the ScanEagle, can fly for 25hrs with a 2 stoke engine.
RCAV8R13

[kd7ost]

I was just through your town a month ago. My in-laws are just across the river from you in "The Dalles". I was wondering about trying to get a tour of your facility? My in laws have a young neighbor, I don't recall his name, but he's been working there at your facility for a short time on the ground stations. He asked about a tour for me but told me that he had been told family only for now.

Anyhow, were you working for Aerosonde when the Atlantic flight took place?
I met Greg Holland in Boise last year. I work for the BLM at the National Interagency Fire Center and he was out giving a speil on how the Aerosonde could work in the Wildland Fire management scenario.

I build and fly my own version of a UAV but restrict my flying to inside of the "See and avoid" concept. I'm small time and am just turning the hobby into a part time business venture. You can see the plane I designed and built here. http://www.rcapa.net/forums/viewtopic.php?t=285 It's a pretty conventional platform. My capabilities are extremely limited by comparison. I only get 1 hour of flight time on my 32 ounces of gasoline. (Fuji BT32A) I can fly autonomously for that hour though if I program it up and let it go.

What does it take to get the kind of "mileage" you guys get? When I talked to Dr Holland he said that was the single biggest acheivement for them. Fuel injection tied in with avionics was all he would say. The engine isn't a stand alone product I surmise?

Thanks for your info. I've seen the Laima on diplay at the Air Museum up in Seattle. (Where I'm from) Way to go.

Dan

[RCAV8R13]

I was just through your town a month ago. My in-laws are just across the river from you in "The Dalles". I was wondering about trying to get a tour of your facility? My in laws have a young neighbor, I don't recall his name, but he's been working there at your facility for a short time on the ground stations. He asked about a tour for me but told me that he had been told family only for now.

Anyhow, were you working for Aerosonde when the Atlantic flight took place?
I met Greg Holland in Boise last year. I work for the BLM at the National Interagency Fire Center and he was out giving a speil on how the Aerosonde could work in the Wildland Fire management scenario.

I build and fly my own version of a UAV but restrict my flying to inside of the "See and avoid" concept. I'm small time and am just turning the hobby into a part time business venture. You can see the plane I designed and built here. http://www.rcapa.net/forums/viewtopic.php?t=285 It's a pretty conventional platform. My capabilities are extremely limited by comparison. I only get 1 hour of flight time on my 32 ounces of gasoline. (Fuji BT32A) I can fly autonomously for that hour though if I program it up and let it go.

What does it take to get the kind of "mileage" you guys get? When I talked to Dr Holland he said that was the single biggest acheivement for them. Fuel injection tied in with avionics was all he would say. The engine isn't a stand alone product I surmise?

Thanks for your info. I've seen the Laima on diplay at the Air Museum up in Seattle. (Where I'm from) Way to go.

Dan
Hi Dan,
I'm sure I can set up a tour for you. I gave a tour to some of our club members a few months ago.
I have been with The Insitu Group for over 11 years. I've been there since day one. As the test pilot and chief aero tech, I have built and flown them all (Aerosonde, Viator, SeaScan, ScanEagle and every abortion in between). I flew the take off in Newfoundland of the Atlantic crossing. I developed the engine used in the Aerosonde. I invented the unique oiling system that the Aussies still use today. The oiling system uses a set of check valves and positive crankcase ventilaton to circulate the oil in and out of a tank. This way we don't have to dilute the fuel with oil. Add to that a very large, low pitch prop and you burn about 275g per KWH. The original engine was a converted Enya R120 four stroke running 100LL avgas. The Aussies did the fuel injection system that is now used on an Enya R155.
BTW, I'm also from Seattle.
PM me the next time you are in town.
RCA
P.S. I'll check out your UAV.

[sesat]

Wow!

[danstrider]

I'll second the wow!

I saw an Aerosonde in Baltimore at the 2005 AUVSI conference and I was duely impressed. Ironically, nobody staffing the Boeing booth ever came over to answer my questions. I must have looked inapproachable because I was engulfed in the details of boom fairings, engine modificatinons, tail construction, hollow wing construction. The Aerosonde on display had spacers between the wing and fuse pod, looked like for changing wing/tail incidence compared to the fuselage, though it left an unsightly gap at the interface. Other than that (which nobody knew an answer to), it was neat to see Aerosonde in person. I highly recommend travelling if anyone has an option to see it (I'd work on RCA for a tour also, but I'm an east-coaster and WA is a bit of a haul :-)

Anyhow, cool that we know your background now RCA! My compliments to the chef.

Dan

[kd7ost]

Let me try again below. Something isn't posting right here.

Dan

[kd7ost]

Hi Dan,
I'm sure I can set up a tour for you. I gave a tour to some of our club members a few months ago.
I have been with The Insitu Group for over 11 years. I've been there since day one. As the test pilot and chief aero tech, I have built and flown them all (Aerosonde, Viator, SeaScan, ScanEagle and every abortion in between). I flew the take off in Newfoundland of the Atlantic crossing. I developed the engine used in the Aerosonde. I invented the unique oiling system that the Aussies still use today. The oiling system uses a set of check valves and positive crankcase ventilaton to circulate the oil in and out of a tank. This way we don't have to dilute the fuel with oil. Add to that a very large, low pitch prop and you burn about 275g per KWH. The original engine was a converted Enya R120 four stroke running 100LL avgas. The Aussies did the fuel injection system that is now used on an Enya R155.
BTW, I'm also from Seattle.
PM me the next time you are in town.
RCA
P.S. I'll check out your UAV.

Excellent. Thank you.

Your post above said you're now using a 2 cycle engine with the scan eagle and have 25 hour duration.

What can guys like me do, using Fuji and Zenoah gas engines, to increase our range if you don't mind me picking your brain? How much effort is it for the modeler building in his garage and how much can be off the shelf mods? Or, do you use a different approach all together? I get 1 hour for 1 quart of 2 cycle gasoline mix with either a Fuji BT32A or a Zenoah G26.

Thanks

Dan

[clolson]

Excellent. Thank you.
Your post above said you're now using a 2 cycle engine with the scan eagle and have 25 hour duration.
What can guys like me do, using Fuji and Zenoah gas engines, to increase our range if you don't mind me picking your brain? How much effort is it for the modeler building in his garage and how much can be off the shelf mods? Or, do you use a different approach all together? I get 1 hour for 1 quart of 2 cycle gasoline mix with either a Fuji BT32A or a Zenoah G26.
Thanks
Dan

Let me just add a "me too" to Dan's question. I'm sure that long endurance is part of these company's competitive advantage, but it would be very interesting to start collecting some of the tricks and magic online someplace. I bet if someone was selling modified high endurance engines/fuel systems, there would be a market for that sort of thing ...

There's a cascading effect though ... if you are up in the air for 25+ hours you are going to start having serious battery issues ... that implies an onboard generator which increases complexity, cuts into your fuel economy, cuts into your payload, etc. But a generator would let you fly at least as long as the engine runs.

I haven't done any real fuel economy tests on my Kadet Sr. platform, but with a 4-stroke glow, I estimate that I can get maybe 3 minutes per oz of fuel when I'm throttled back and barely holding altitude. If I could figure out places to put a gallon of fuel ... that might give me 4+ hours of flight time (assuming I'm going to have to burn fuel faster at the beginning of the flight when I'm heavier.) By my calculations a Kadet Sr. could carry the weight of a gallon of fuel, but when I set the jug next to the airplane I end up scratching my head for where to put it all ... and then I start wondering if my little engine would be able to consistantly suck the fuel out of that large of a tank system all the way from full to empty ... (?)

Then I start thinking that if you have a UAV up for multiple hours, part of that time might be spent loitering, waiting for something interesting to happen, part of that might be spent flying pedal to the metal to get to an interesting even as soon as possible ... different throttle settings at different times. That makes it extremely difficult to estimate fuel remaining. You would almost need some sort of sensor to report fuel levels, or at least a trigger/warning for when you drop below some threshold so you know it's time to head for home or the nearest landing site.

Curt.

[ockie]

I've been playing around with the idea of a privately built long range UAV for quite some time now. In the end it all seems to boil down to one thing: if you want good range you need a darn efficient powerplant. All the rest (autopilot, aircraft, telemetry, etc.) is "reasonably" straight forward.

One interesting idea is a diesel conversion on a standard 4-stroke glow engine (http://www.rcnightflying.com/nfimages9.html), which apparently has excellent fuel efficiency. Maybe one of the new gasoline TRUE 4-stroke (sump with oil!) aero engines by Fuji (BF25/BF34), or even an aero converted Honda 4-stroke (GX25/GX31) (http://www.carrprecision.com/Pages/prod02.htm) could be tweaked for better performance.

However, here's my favourite idea (please don't laugh): what about a LPG conversion on one of the above mentioned Fuji / Honda engines? It will have to be a true 4-stroke engine with non fuel based lubrication (sump & oil) for a LPG conversion to work, I suppose. LPG is a gas allready and on that basis alone should combust better than the "fuel-mist" that our small scale carburetors produce. Maybe its excellent combustion behaviour far out weighs its lower energy content than normal glow fuel / gasoline.

Is there anyone out there who thinks this might be worth investigating?

[gcouger]

If you can get the tank weight down the efficiency is better than you think. Propane weighs 4 pounds to the gallon and gasoline weighs 7. A properly set up propane engine will get the same mileage as gasoline. You have to raise the compression ratio to 15 to 18 to 1 to fully realize the potential of propane. Ideally direct injection gets the best out of any fuel but fuel that are already gas are not as critical.

Iso-butane might be a better choice of fuels. The boiling point is much higher so the tank pressure is much lower than the 260 psi of propane on a hot day. The down side is you will have to heat the iso-butate to boil off the vapor with some kind of heat exchanger where you might not have to with propane.

Another source of inefficacy is the mixture often has to be run rich to keep from burning too hot and eroding the exhaust valves. I never knew anyone to run an engine on Iso-butane so the problems may be different.

I have seen a FAA approved propane conversion for a private plane back in the 50's.

Gordon Couger
Stillwater, OK
www.couger.com/gcouger

[gcouger]

Alan made the entire plane and most all the electronics for it also. He even designed a new brushless controller that far out performs the usual ones that are used for most model applications. While most controllers are pretty good at full or near full power the system efficiency falls off very fast as the power levels drop. In order to have a chance at the 24 hour endurance he needed something better and he set out to build it and did. the plane used a Kontronik Tango motor and a 9 phase controller. At full power the system efficiency is over 90%--at 10% power it declines less than 5%. Operating at the same power with "normal" controllers would yield a reduction of efficiency on the order of 40% or more.

The solar cells were molded into the upper surface of the wing. I have some more pictures that I can post later.

I did a few hours of flying for Alan during the early afternoon of the 21st of April. It was a treat to fly thermals from video with a full set of flight instruments and GPS.

Steve

Hi Steve

How do I get in touch with Alan? I am working the same problem with running the motor at low speed. Binning in with problems that are already solved would be great.

Gordon
Gordon Couger
Stillwater, OK
www.couger.com/gcouger

[sesat]

Can't we easily get 40 miles or more with electric sailplanes, discounting thermal activity?

2500m of climb * l/d ratio > 25 = 40mi range...

... i think...


Ram.

[RCAV8R13]

Dan, Curt and others,
Sorry it took soo long to chime in but here are my thoughts on fuel efficient engines.
We use the 3W-24i on the ScanEagle. 24cc with a 16x14 APC prop turning 6,200 rpms with everything turned on (battery charger, video system etc). To get better fuel economy we replaced the large Walbro carb with a smaller Walbro WT456 carb. This cost a little performance but greatly decreased fuel consumption. We set the whole plane in a test stand and tune the carb as lean as we can with a full load on the engine. We measure fuel consumption at 10 different throttle settings. We load this data into the avionics and use it to estimate fuel remaining during the flight. We also had Dubb Jett of Jett engineering design and build us a very special version of his Turbo Jett muffler that makes the engine very quiet while only pulling 200-300 rpms off the top end, compared to a straight header pipe. We run 92 octane pump gas and set the timing at 23 deg BTDC. That's about it.
On the Aerosonde, we used the Enya R120 4 stroke with the same Walbro WT456 carb running on 100LL avgas. But we used an APC 20x8 prop. We did not mix the fuel and oil, however. We used an oiling system of my design that uses positive crankcase ventilation and a series of check valves to circulate oil through the engine and in and out of a tank (which is ventilated). This was key to getting across the Atlantic as we burned only 30cc of oil during the flight While the fuel remained undiluted by oil. Curious, Honda's new little 4 stroke uses the very same concept!
Other than that, all you need is lots of wing area, low drag and lots and lots of fuel.
RCA

[kd7ost]

Dan, Curt and others,
Sorry it took soo long to chime in but here are my thoughts on fuel efficient engines.
We use the 3W-24i on the ScanEagle. 24cc with a 16x14 APC prop turning 6,200 rpms with everything turned on (battery charger, video system etc). To get better fuel economy we replaced the large Walbro carb with a smaller Walbro WT456 carb. This cost a little performance but greatly decreased fuel consumption. We set the whole plane in a test stand and tune the carb as lean as we can with a full load on the engine. We measure fuel consumption at 10 different throttle settings. We load this data into the avionics and use it to estimate fuel remaining during the flight. We also had Dubb Jett of Jett engineering design and build us a very special version of his Turbo Jett muffler that makes the engine very quiet while only pulling 200-300 rpms off the top end, compared to a straight header pipe. We run 92 octane pump gas and set the timing at 23 deg BTDC. That's about it.
RCA

Cool, Thanks RCA, I just figured you were out flying someplace very dry or very wet.

Does that smaller Walbro carb bolt right on in place of the larger one or do you have to make an adapter? I have a WT645 on the little Zenoah and a WT407 on the Fuji. I haven't tried or even considered trying to swap them out but they sure look like they are interchangeable. The bolt pattern seems to line right up if I measure it.

Dan

[kd7ost]

Cool, Thanks RCA, I just figured you were out flying someplace very dry or very wet.

Does that smaller Walbro carb bolt right on in place of the larger one or do you have to make an adapter? I have a WT645 on the little Zenoah and a WT407 on the Fuji. I haven't tried or even considered trying to swap them out but they sure look like they are interchangeable. The bolt pattern seems to line right up if I measure it.

Dan


RCA,

I answered some of my own question by fumbling through the www.walbro.com site. It looks like all the WT carbs are physically the same size externally for mounting. When you try to get more details though, there isn't much there. I can't even find a listing of the WT carbs with any specs. Do you have a source or link that shows more than the walbro site does?

Thanks again

Dan

[RCAV8R13]

RCA,

I answered some of my own question by fumbling through the www.walbro.com site. It looks like all the WT carbs are physically the same size externally for mounting. When you try to get more details though, there isn't much there. I can't even find a listing of the WT carbs with any specs. Do you have a source or link that shows more than the walbro site does?

Thanks again

Dan
You could try calling Lois Salas at Walbro's tech line. I don't have the number here at home but I do have it at work. The number could be on the website. If you can't find it, I'll find it and post it here. Lois can give you all the help you need, tell him "Kip Jackson of The Insitu Group" sent you.
RCA

[hugo_vincent]

Hi RCA,

Can you tell us anything about the fuel injection system used on the newer Aerosondes?

I have been toying with the idea of building a digital fuel injector using an inkjet assembly out of a printer... need to check material compatibility with fuel etc. and figure out if I can deliver enough fuel to the engine fast enough. But it seems driving the inkjets themselves is not too hard (I am an elec. engineer).

Cheers,
Hugo.

[djklein21]

fuel is very corrosive, I doubt the ink jet drivers will work

[ockie]

Will that fuel injection system be used on the Albatros then, Hugo? :-)

What's the status on your project, have you done a final report yet? ...would love to give it a read...

-Ockert

[hugo_vincent]

Hi Ockert! Long time, no hear...

Yes, thats the plan if it works!

I have PMed you about the final report.

Also, check the website at: http://www.albatross-uav.org

Fuel corrosiveness: yes, fuel is very corrosive... but inkjet inks are also highly specialized and quite corrosive in their own way. From what I can gather so far, most inkjets are made from a silicon chip (with the actual inkjets on it) covered in a protective coating, a glass jet plate and ink reservoir, and a fiberglass, FFC and/or aluminium substrate. I pulled apart an old inkjet cartridge and separated out the actual inkjet assembly, and left it to soak in glow fuel overnight, and there doesn't seem to be any visible material damage... need to try it in gasoline next.

[blimpyway]

Noticed sometimes ago two strange, unrelated ideas. Dont remember any links..

First one is a flock of gliders. Each of them an UAV by itself communicating somehow (W-LAN?) with each other. When any of them notices a thermal it broadcast its position to the rest of the folck, and all rush in to gain altitude. Once high enough they spread towards the desired destination each seeking the next thermal.

----------

The second one is really weird: jetstream riding with a kite. How it works - a payload is attached to a lightweight kite with a very long thin line - one mile or more. The payload has a steerable parachute for drag, (or even beter a gyro propeller to generate power too). The point is to exploit wind gradients between different altitudes. Launch the whole assembly with a balloon., with the whole 1mile long line stretched. When the balloon senses wind changed radically within last 1/2 mile up, it releases the line. The "aircraft" will probably keep drifting as long as there-s enough wind gradient.

There-s also a wind gradient close to the surface - between very low (20-50 meters) altitude and higher altitude (200-400meter), wind turbines get significantly more wind as the tower is higher. So during a windy day, over large, unpopulated flat terain it might stay aloft for a long ride.
The steerable drag chute can be used to seek better ascendence.
And dont expect to recover the "aircraft" .

cezar

P.S. Here, (down within the page) is described a dual kite flight method: http://www.yellowairplane.com/MISC/Tethered_Airfoils/

[charlieandrews]

Dan, Curt and others,
Sorry it took soo long to chime in but here are my thoughts on fuel efficient engines.
We use the 3W-24i on the ScanEagle. 24cc with a 16x14 APC prop turning 6,200 rpms with everything turned on (battery charger, video system etc). To get better fuel economy we replaced the large Walbro carb with a smaller Walbro WT456 carb. This cost a little performance but greatly decreased fuel consumption. We set the whole plane in a test stand and tune the carb as lean as we can with a full load on the engine. We measure fuel consumption at 10 different throttle settings. We load this data into the avionics and use it to estimate fuel remaining during the flight. We also had Dubb Jett of Jett engineering design and build us a very special version of his Turbo Jett muffler that makes the engine very quiet while only pulling 200-300 rpms off the top end, compared to a straight header pipe. We run 92 octane pump gas and set the timing at 23 deg BTDC. That's about it.
On the Aerosonde, we used the Enya R120 4 stroke with the same Walbro WT456 carb running on 100LL avgas. But we used an APC 20x8 prop. We did not mix the fuel and oil, however. We used an oiling system of my design that uses positive crankcase ventilation and a series of check valves to circulate oil through the engine and in and out of a tank (which is ventilated). This was key to getting across the Atlantic as we burned only 30cc of oil during the flight While the fuel remained undiluted by oil. Curious, Honda's new little 4 stroke uses the very same concept!
Other than that, all you need is lots of wing area, low drag and lots and lots of fuel.
RCA

RCA,
Just out of curiosity, do you know what the wing area of the ScanEagle is off the top of your head?

Cheers,
Charlie

[RCAV8R13]

RCA,
Just out of curiosity, do you know what the wing area of the ScanEagle is off the top of your head?

Cheers,
Charlie
Charlie,
It's about 800 sq in.
RCA

[wadiprawita]

to RCAV8R13,

I have 3W 28i, and want to mount a AXI 2212-34 as the generator ( I have used this in zenoah 26, works well on the bench, haven't tested it in flight though). But I found that it's quite difficult to make the generator coupler with this 3W 28i.

Any recommendation?

What motor do you use as the generator?

And sorry, I also posted this in other thread, before found this thread.

Thank's

-doni-

[RCAV8R13]

to RCAV8R13,

I have 3W 28i, and want to mount a AXI 2212-34 as the generator ( I have used this in zenoah 26, works well on the bench, haven't tested it in flight though). But I found that it's quite difficult to make the generator coupler with this 3W 28i.

Any recommendation?

What motor do you use as the generator?

And sorry, I also posted this in other thread, before found this thread.

Thank's

-doni-
Doni, PMed you.

[keertib]

Dan, Curt and others,
Sorry it took soo long to chime in but here are my thoughts on fuel efficient engines.
We use the 3W-24i on the ScanEagle. 24cc with a 16x14 APC prop turning 6,200 rpms with everything turned on (battery charger, video system etc). To get better fuel economy we replaced the large Walbro carb with a smaller Walbro WT456 carb. This cost a little performance but greatly decreased fuel consumption. We set the whole plane in a test stand and tune the carb as lean as we can with a full load on the engine. We measure fuel consumption at 10 different throttle settings. We load this data into the avionics and use it to estimate fuel remaining during the flight.

...<snip>...

RCA


This is awesome information! Thanks a lot to everyone on this thread! I am looking to setup a similar system to estimate fuel consumption for my UIUC UAV group, and I was wondering if there was any literature or links that would better help me understand how to correlate RPM to fuel consumption. If I understand it correctly, the fuel to air mixture is not exactly constant even at a constant RPM (i.e. the engine uses more gas when trying to climb vs. diving at the same throttle setting), but it is greatest during on the ground static-testing. Any references on how fuel consumption and/or carbs work on these engines would be great. I'm going to be trying to predict fuel left using an RPM sensor for a O.S. 1.60 FX and Brillelli 46cc.

One other thing I was wondering about...fuel gauges? Why are they not used in many UAVs? Cost? Size? Availability?

Thank you,
Keerti

[Connexxion]

Don't forget one thing:

The mixture changes as you climb.Actually a modified altimeter coupled to the mixture screw would be perfect.During the climb the mixture will be adapted to the altitude the engine is running on.
Higher (less oxygen) = leaner and lower (more oxygen)= richer

[Mark Harris]

fuel gauges are usually made with flow meters.. they are very expensive.

[RCAV8R13]

Don't forget one thing:

The mixture changes as you climb.Actually a modified altimeter coupled to the mixture screw would be perfect.During the climb the mixture will be adapted to the altitude the engine is running on.
Higher (less oxygen) = leaner and lower (more oxygen)= richer
Active mixture control is not necessary. Walbro and Tilotson carburetors have fuel pressure regulators that use ambient air for their reference. We have flown to 20,000 ft without touching the needles. The carburetor compensates automatically.
RCA

[wadiprawita]

Dear RCA,

do you use factory modified 3w pusher version with standard prop, or standard engine with pusher prop ?

-doni-

[Connexxion]

Active mixture control is not necessary. Walbro and Tilotson carburetors have fuel pressure regulators that use ambient air for their reference. We have flown to 20,000 ft without touching the needles. The carburetor compensates automatically.
RCA


Could you provide us a link to system you mentioned?

[kd7ost]

Walbro and Tilotson carbs are the ones that come stock on the gasoline engines you buy. Not just for RC through Fuji, Zenoah, 3W and the like. But even gas powered weed eaters, leaf blowers, lawn mowers etc.

Dan

[RCAV8R13]

Dear RCA,

do you use factory modified 3w pusher version with standard prop, or standard engine with pusher prop ?

-doni-
We reverse the engine and run off the shelf tractor props.
RCA

[Earl Jr.]

Regarding telemetry for a endurance UAV; one of the ideas I have herd thrown about is using a satellite phone as a modem. It likely wouldn't provide the bandwidth necessary for all the data, but would have enough to provide support for basic navigation. I'm not attempting to pimp any particular carrier, but Iridium has a satellite modem (http://www.bluecosmo.com/shop/product/71/Globalstar_GSP-1620_Satellite_Data_Modem/) that could easily be adapted. I'm sure there are other carriers out there more suited to this. One concern I have about satellite data (other than cost) is the inherent latency. I hope your autopilot is exceptionally robust. :)
I don't have experience with UAV's specifically, but I live in an exceptionally mountainous area that limits your range unless you have quite a bit of altitude, hence having to entertain alternative data links.

[chrisgornall01]

I'm playing with an interesting idea for long range telemetry, same way as I am typing this, my laptop is plugged into my mobile phone which supports using 3g network for high speed internet access.
Looking at using a nano itx based system, and programming a simple directplay app (as used in multiplayer games over the internet) for high speed bi directional comms to base station, by high speed i mean max data rate of 112kbps, (network can support more, limiting factor is not wanting to buy new phone).
As for cost, my phone provider do unlimited internet access for as little as £7.50 a month!

[joelhaasnoot]

Wouldn't it be better to use the phone as a modem and use TCP/IP? DirectPlay I guesds also uses TCP/IP, but can also use IPX i believe, which is pretty old...

[chrisgornall01]

Thats what i'm doing joelhaasnoot, only using the 3g net to connect via packet switched (digital), gives much improved transfer rate, my intention is to (eventually) stream video + telemetry and send control signals from base station joystick.
Autonomous flight for waypoints and connection loss, with the option of direct control when required.

[Earl Jr.]

The reason I was focusing on satellite instead of cellular for communications is because cell coverage in southcenteral Alaska is spotty and there is a lot of water and mountains where I would be flying.

[macboffin]

A motor to charge a battery is using fuel to do so; the fuel is potential energy.You will need to use some battery power to start the engine; there will be energy losses due to thermal efficiency, friction etc in engine and drive train to generator and thence to battery via voltage regulator etc; and further losses from battery to motor etc.Total efficency of the system likely to be somewhere around twenty per cent if you are lucky!
Better to design a low drag aircraft, ie a powered glider which climbs to altitude, powers up whenever it gets low, or can maintain altitude on very low power. Bear in mind that since pressure drops with altitude, so does engine power, and you need automatic mixture control to maintain stoichometric balance, otherwise the mixture gets progressively richer the higher you go. Maximum duration is got at lowest altitude and slowest practical speed; drag and thus power requirement goes up as the square of the velocity; ie twice speed requires four times power; doubling the power gets you a speed increase of only twenty five per cent.(Man -powered aircraft typically fly at eight to ten mph since only about half a horsepower is available continously). Apart from all this, and auto nav, whether GPS or other system, what about "other traffic" avoidance?

[picooZer]

I'm new to all this UAV stuff, in fact im new to rc groups.

One thing I've been very curious about is the endurance of rc jet planes. The stats of planes such as the global hawk are impressive: 42 hours of flight time at 450 miles per hour is impressive. But just how efficient is a rc jet? Might start a turboprop project in the next ten years.

[danstrider]

Hi PicooZer,

RC Jets are pretty inefficient compared to the equivalent IC and Electric counterparts. Remember that jets get more efficient as speeds increase because the incoming air is pre-compressed by the forward speed rather than having the engine's compressor stage(s) doing all the work. Also, model turbines are typically a single stage centrifugal compressor rather than a several stage axial compressor that would be found on the Global Hawk.

On the RC size scale, a jet project I worked on with a 15lb thrust Mercury turbine carried approximately 1/2 gallon of fuel for 11 minutes of full throttle run time. Only 3/4 gallon of fuel was used by an IC engine to fly 1,900 mi across the Atlantic (http://www.sciencenewsforkids.org/articles/20031217/Feature1.asp) (http://www.sciencenewsforkids.org/articles/20031217/Feature1.asp%29). Starting down the turboprop line might be a higher efficiency solution if you're set on using a jet. It's more efficient to accelerate a lot of air a little than to accelerate a little air a whole lot.

Dan

[jetjock51]

I'm new to all this UAV stuff, in fact im new to rc groups.

One thing I've been very curious about is the endurance of rc jet planes. The stats of planes such as the global hawk are impressive: 42 hours of flight time at 450 miles per hour is impressive. But just how efficient is a rc jet? Might start a turboprop project in the next ten years.

I have been flying turbine jets since 1998. The most common class of turbine in use today puts out 27 to 30 lbs of thrust and burns approx. 12 oz of Jet-A or Kero per minute. For UAV purposes, they are very inefficient. However, they sure are a lot of fun!