Solar Powered UAV Project

[MIT KID]

Hi Everyone,

My name is Adam and I'm a senior in the Aero/Astro department at MIT. My partner and I are designing/building/testing a solar powered UAV for our senior capstone course, Experimental Projects I/II. It is a two semester course consisting of pairs of students who design and carry out their own experimental project. Last semester was dedicated to the design of our aircraft and the associated testing scheme. This semester we will actually build the aircraft, and test it to se how closely its performance matches the various computer models we've developed. I figured I'd start a build log for the project, so that I'll be able to remember where a year of my life went :)

The course administrators place a $500 budget cap on each project, and since we have essentially all the supplies inhouse we needed to put the vehicle together, we decided we would use the $500 mark as a cap on our Solar Array cost. Since we were going to devote so much time to the project, we decided to make the aircraft meet the FAI f5sol requirments, so we could qualify for any duration records we might set along the way (current record is 11+hrs). As a result the design constraints were as follows.

$500 array limit (we sized it for $400 so we could replace broken cells if needed)
900g minimum weight limit
0.75m^2 projected wing + stab area
3.5m span limit (for transportation and fabrication reasons)

We wrote a multidisciplinary optimization script in MATLAB that maximized power margin (details for those who are interested) to come up with an optimized wing configuration, then we selected motor/number of cells/stab areas to match the wing.

Our Final A/C configuration is as follows:

S=0.72m^2
B=3.31m
AR=13.6
Weight~10N
Airfoil=NG01
Cruise Thrust= 0.33N @ 5m/s

The Array will consist of 18 A-300 Sunpower Solar Cells (we are still looking for a supplier if anyone has any ideas), which will put out about 56watts at full power. There will be an pic controller onboard that serves as a maximum power point tracker and automatically regulates motor power to maximize the array's efficiency.

The generous people at Maxon Motors have provided us with 90% brushed motors and Polecat Aero has supplied us with EZ bubble dancer stab parts, which will make construction a bit easier.

I guess thats about it for now. We finished cutting wing cores today and will start layup later this week. Questions and discussion are welcome. Pics will follow.

-Adam

[kd7ost]

Good stuff Adam,

Although I don't know a source for the panels, I do recall some threads in the past that might be helpful to you. The guys that did the 48 hour endurance flight have posted on here. One looks pretty prolific. You might concider contacting them via email if they allow it or at least PM for some information on purchasing the cells. They used the same ones you're looking for.

Please keep posting here. It sounds fun and we would like to watch. ;)

Dan

http://www.rcgroups.com/forums/showthread.php?t=378861

http://www.acpropulsion.com/

http://www.acpropulsion.com/ACP_PDFs/AC_Propulsion_Solong_UAV.pdf#search=%22A-300%20Sunpower%20Solar%20Cells%22

[MIT KID]

Good stuff Adam,

Although I don't know a source for the panels, I do recall some threads in the past that might be helpful to you. The guys that did the 48 hour endurance flight have posted on here. One looks pretty prolific. You might concider contacting them via email if they allow it or at least PM for some information on purchasing the cells. They used the same ones you're looking for.

Please keep posting here. It sounds fun and we would like to watch. ;)

Dan

http://www.rcgroups.com/forums/showthread.php?t=378861

http://www.acpropulsion.com/

http://www.acpropulsion.com/ACP_PDFs/AC_Propulsion_Solong_UAV.pdf#search=%22A-300%20Sunpower%20Solar%20Cells%22

Thanks for the heads up on those threads. The so-long 48 hour guys used A-300's on their aircraft. We were going to ask them about where they purchased their cells, nice to see some of them are fellow rcgroupers :) .

We tried to get cells from the supplier who provides a-300 for the MIT solar car team, but they have an agreement with sunpower to sell only encapsulated cells, and we need just the bare cells.

Anyways I'll be sending out a bunch of pm's and see if anyone's got a line on A-300's

[kd7ost]

We tried to get cells from the supplier who provides a-300 for the MIT solar car team, but they have an agreement with sunpower to sell only encapsulated cells, and we need just the bare cells.

Bummer,

I take it you've tried to get them direct from sunpower after explaining your project and the need for getting them without the packaging?

http://www.sunpowercorp.com/solarpanels/

Dan

[Myron]

Hey MIT Kid,

We tinkered with the idea of using some solar power in one of our birds.. We found an outfit here in Austin that is doing some incredible stuff with solar power. They have a coating that is kinda painted on. Its a little pricey at 300.00 per watt but I am sure the price will come down as technology gets better... Google them up at Heliovolt...

Myron

[rclinks2002]

Yay for Matlab!!!! catia is sweet also

[MIT KID]

Yay for Matlab!!!! catia is sweet also

HAHA yah. Whenever I click on my start bar Matlab is at the top of the most used programs...kind of sad :)

[MIT KID]

So after about 6hrs of cutting, all the cores are done. We did all the cutting on the department's CNC hot wire. It's notoriosly finicky (all the error messages are in french) and hadn't been used since last semester, so it took about 8 practice cuts to get everything working smoothly. We were a bit worried about cutting the tips because the airfoil is only about 5.5% thick and we have a tip cord of only 10cm, but the everything came out really good. We did the tip pannel in two sections to prevent melting, so we've ended up with a 7 pannel wing :eek:
Core prep is scheduled for tomorrow, and we should be ready to start laying things up this weekend.

BTW we used Dow Highload-60 foam for everything

[Tophinater]

This isn't fair. You guys are doing what I wanted to do for my senior project. :(

Anyways, what are you guys going to be using for a power plant? Also what do you expect your total solar wattage to be in direct sunlight? And what equipment are you guys going to be using and how much power do you expect it to consume?

[CAFplanekid]

Awesome project, I am lookin forward to seeing it come together.

I remember seeing your plane spiral in, that was an awesome job getting it flying again for the next day.

[MIT KID]

This isn't fair. You guys are doing what I wanted to do for my senior project. :(

Anyways, what are you guys going to be using for a power plant? Also what do you expect your total solar wattage to be in direct sunlight? And what equipment are you guys going to be using and how much power do you expect it to consume?

here's our motor:

http://www.maxonmotorusa.com/files/catalog/2005/pdf/05_077_e.pdf

Our open circut array voltage is 12v but we'll be opperating at about 70%, which is where the cells are at max eff. Our array's maximum array wattage is 55watts, but the aircraft is specked to cruise at about 4watts. This provides us with a power margin of 13.75, which will allow us to sustain flight when the sun is only about 6.5 degrees above the horizon. Our motor is rated for 15watts which will provide a climb rate of 140 ft/min, not spectacular, but we had to compromise to max out the propulsion systems eff.

[MIT KID]

Today we readied th cores for bagging. We cut the polyhedral angles, trimmed the TE's, smoothed out the LE's, and cut all of the spar pockets. I also started work on the wing joiner blocks, and did a bit of work on the pod.

we have a 3 day weekend so I'll be putting in some long hours on the project.

[MIT KID]

...but they are made out of paper :p . Kidding aside, we have finally found a supplier who can get us A-300's. He will wire and encapsulate the cells, and then we will attach the panels to the wing. Today I cut out 1:1 scale cells to finalize the layout and come up with plans for our cell supplier to work from.

The encapsulated panels will only be about 1mm thick so we shouldn't have any problems with them doing bad things to the airflow. Prof. Drela said that the NG01 airfoil we're using likes to have a turbulator near the LE so the step created by mounting the panels ontop of the glass might actually help out.

As far as wing progress goes, we're almost done with the prep for glassing, all thats left to do is mount the joiner blocks and install wing bolt hardpoints.

[MIT KID]

Here's a schematic of our maximum power point tracker. Now back to writing my stupid paper for "History and Theory of Architecture," I really dislike the humanities requirement :mad:

[MIT KID]

the plane is starting to come together. Tips are finished, and the center section is ready for bagging tomorrow (we've prebagged the hardpoints, spars, and unicaps to make the actual layup easier). Fuselage bulkheads are also done, and the stabs are framed up and ready for sanding. Its a 4day wekend, so Hopefully the airframe will be done, or almost done by wednesday.

[IneptOne]

Hi Adam,

I think you might need to do a little more work on your maximum power point tracker. The way it is configured now, you will have pulsed current from the solar cells unless your PWM frequency is very high.

Unless you are planning to use a PWM frequency on the order of 100kHz or greater I think your filter components are too small.

Dan

[saucerguy]

I guess the goal for breaking the endurance flight record is to get enough extra juice to trickle charge enough back up batteries to allow it to operate without any sun at all, enough to keep it airborn overnight, if achieved, you'd be flying it for an unlimited amount of time. That bird looks very graceful, can't wait to see the final project, also look at envy with your foam cutting set up, I'm still doing it with a manual bow.

[MIT KID]

Today we took the center section out of the bag, and the results were awesome :D . We were kind of worried about bagging such a large and complex section as one piece, but every thing seems to have stayed in place. Now that all of our major sub assemblies are done, we'll start finish work and final assembly.

IneptOne- thanks for the heads up on the possible problem with the MPPT, I'm going to have one of the electronics guru's in the department take a look at our design.

saucerguy- Thanks for the kind words. There wont actually be a battery on board. We're going for the FAI F5SOL duration record which among a bunch of other constraints says that there cannot be a battery buffer on-board, just Solar. So... we optimized the design for maximum power margin (power required/power available) this will allow the plane to stay aloft at extremely low sun angles, letting us launch earlier and stay up longer.

[libelulamodelos]

hey cool project , congratulations !!

[hemna]

nice job guys. did you create the mold for the fuse as well?
what vaccum system did you use?

[JSAMVTOL]

Interesting project. I see someone has already queued in to the So Long from AC Propulsion.

Curious though:

1. Will you charge conventionally before flight? or leave the plane out in the sunlight for charge, or attempt to re-charge from the array in-flight to prolong flight.

2. SpectraLabs in LA makes some cutting edege solar cells ($$$), touting 40%+ efficiency for a superamorphous type cell. They dope the silica differently that most conventional store bought cells. Instead of spreading out the array, I've seen some where cells are dense (more depth) to minimize area use without loss of efficiency + power.

3. Do you plan on following thermals to enhance flight and allow cells to re-charge in flight?

4. One suggestion to combat radiation, you may have lots of it. Use a windturbine (2) mounted on both side with ducts for cooling aire. Or even better use the wind turbine in-flight to re-power the batteries...supplement the array. I'm trying that with my scratch build....but watch out for drag penalties.

Good Luck. Let us know if it works.

[MIT KID]

Interesting project. I see someone has already queued in to the So Long from AC Propulsion.

Curious though:

1. Will you charge conventionally before flight? or leave the plane out in the sunlight for charge, or attempt to re-charge from the array in-flight to prolong flight.

2. SpectraLabs in LA makes some cutting edege solar cells ($$$), touting 40%+ efficiency for a superamorphous type cell. They dope the silica differently that most conventional store bought cells. Instead of spreading out the array, I've seen some where cells are dense (more depth) to minimize area use without loss of efficiency + power.

3. Do you plan on following thermals to enhance flight and allow cells to re-charge in flight?

4. One suggestion to combat radiation, you may have lots of it. Use a windturbine (2) mounted on both side with ducts for cooling aire. Or even better use the wind turbine in-flight to re-power the batteries...supplement the array. I'm trying that with my scratch build....but watch out for drag penalties.

Good Luck. Let us know if it works.

Thanks for your questions. Here are some answers....I hope.

1. Again, there will be no propulsion battery onboard. Just the array running throught the MPPT and into the motor. We will have a separate receiver pack whcih we may keep topped off using left over power from the array.

2. We were some what constrained budget wise on the cells. We chose the Sunpower A-300 cells because they offer the most watts per $, are extremely light, and semi flexable.

3. They class requires us to do the whole hypothesis/experiment/measurable outcomes thing, not just say we want to build a solar plane. Consiquently our experiment will be to measure what minimum sun inclination is necessary to sustain level flight. To collect data we will launch about an hour before our predicted minimum necessary inclination and then fly until we can't fly any more, recording the various power outputs and corresponding sun inclination angles. For the record attempts we will surely exploit every hint of lift we can find, climbing as high as we can while the array is putting out the most power.

4. Efficiency loses due to cell heating aren't really a concern for us. The array has been sized to put out enough power that the aircraft can still maintain level flight at very low sun angles (<6 degrees). When the sun is high, and cell heating occurs the array will already be putting out far more power than is needed to fly, so any efficiency hit on the actual cells won't really hurt us. When the sun is low, and we need the array to be opperating at max eff. , cell heating will be negligible.

Answers to hemna's questions:

The molds and fuse were provided by Prof. Drela. They're CNC milled alluminum we still need to mold a conopy hatch, but since the fuse is circular in cross section we can just use the bottom of the fuse mold to make a hatch.

We had to make a special bag, since the center section was so big (the bag was about 8'x3') we made the bag using a large sheet of plastic folded onto itself and sealed aroing the three edges with rubber sealant tape. We used a lot of thick breather material along the TE to make sure we got an even vacuum, and we carried the thin breather cloth almost all the way to the LE. We keept it in the bag for 24 hrs, then removed the mylars and let it sit for an aditional 24 hrs before we started clean up work on the wing.

[MIT KID]

Our motors arrived from maxon yesterday, and they are absolutely beautiful. They're 90% eff. brushed motors, and despite their large size, only put out about 15 watts. We did a trade study, and the added efficiency greatly outweighed the extra weight.

On another note, we're essentially done fabricating stuff. Once I cover the vertical stab, we'll be ready to start final assembly. YAY!

[CAFplanekid]

Wow, beautiful work. You are quite a craftsman. How much do those motors weigh?

[MIT KID]

Thanks, a bunch. You should see Prof. Drela's plane's, everything is immaculate.

the motors weigh a bit under 3oz. Which is coincidentally about the same weight as the recommended rx pack weight for the stock Bubble Dancer.

[CAFplanekid]

I can only imagine what his work would look like. I appreciate craftsmanship more than anything in this hobby, I would rather see a beautifully built rubberband powered model than a giant scale electric with shoddy construction.

I think I remember reading somewhere that brushed motors are more efficient at low RPMs than brushless motors, more or less. Is that what you found and why you went brushed?

[MIT KID]

Exactly. While the brushless motors in the 30watt range claimed 80% eff. and were about 1/4 the weight, at the low rpm we're running (<700rpm) we were getting eff's of only about 45-50%, combine that with a 95% eff controller and things start looking dismal.

[MIT KID]

So, we finally finished all of the parts for the aircraft :D . We dry fitted everything today to make sure our CG was where we want it before we perminantly place the wing saddle. It looks pretty sweet.

Any suggestions on pod color. Should we leave it the natural kevlar or give it a light airbrushing

[tnavressdog]

You have to get some video up of this thing flying! Wow.

[CAFplanekid]

I think you should color it red or black, but I just don't like the look of bare kevlar. Looking forward to a flight report!

[Hep]

Great!

I'd leave the pod natural - composites are already beautiful, but I'm a DLG guy and I like it when the Kevlar plane gets old and turns that ugly mustard color.

What gear are you using for Rudder/Elevator/Speed Control/Receiver?

What are you predicting as the pitch speed and optimal level flight speeds?

I met Dr. Drela out in Poway this year - you're very fortunate. Keep up the good work.

Regards - Hep

[MIT KID]

Great!

I'd leave the pod natural - composites are already beautiful, but I'm a DLG guy and I like it when the Kevlar plane gets old and turns that ugly mustard color.

What gear are you using for Rudder/Elevator/Speed Control/Receiver?

What are you predicting as the pitch speed and optimal level flight speeds?

I met Dr. Drela out in Poway this year - you're very fortunate. Keep up the good work.

Regards - Hep

We're using Hs-65's for rudder/elevator and berg 4L receiver. For battery powered flight tests we'll be using a Castle Creations Pixie-7P, but for solar powered flights we'll be using our home brewed power point tracker.

We came in slightly over our design weight so our new optimal cruise speed will be 5.5 m/s.

Our array came in yesterday, and is absolutely beautiful :D . I forgot to bring my camera with me today, but I'll try and get some pics up tomorrow. Maiden flight isn't too far away.

[MIT KID]

So here are some pictures of the assembled aircraft with our new laminate panels. The panels are a bit under 1/32" thick and their combined weight is 350g. As you can see they are flexible, and lay flat on the wing under their own weight. We will soft mount them to the wing with blenderm tape to prevent cell buckling during wing flex. This was our final cg check before we mount the electrical components.

[MIT KID]

I fininshed up the fuse this afternoon. So now all thats left to do is mold a canopy hatch, and do some final wiring on the panel. Looks like we'll be ready for the maiden flight on tuesday, weather permitting ;) .

The motor mount is three rings, made from 1/16" balsa sandwiched between 1/32" ply. The front ring takes torque loads, the middle ring is loacated at the step where the gearbox meets the motor to take thrust loads, and the back ring is just to assure proper alignment. The motor can be slid out for maintenance or replacement

[rroback]

I built a solar powered model as a senior high school project a few years ago, and face some difficulty in cell choice. I'm not familiar with your cells, but I went with spectrolabs. At thate stage, I wasn't capable of designing my own airframe, so I went with a modified built up hand launch glider. In my calculations I couldn't even create a potentially flight capable model with most solar cells. the spectrolab cells really seem to be the best darn cells out there, the lightest, and most efficient. You airframe seems great, but I'm not sure about the cells. I mounted mine between two layers of mylar, as I didn't want to bond them rigidly to anything. Not sure if my statement help a lot, but I wish you luck. My thing flew, slowly, and barely, but it flew. I wish you more succes

Rhett

BTW, I'm on UCI's d/b/f team. I loved seeing your big plane try and lift 5 bottles last year.

[MIT KID]

Rhett,

Do you have any pics or your plane? We considered the Ga As cells from spectrolabs (they're the ones that the solar car team uses) but they were prohibitively expensive for this project.

The sunpower cells are pretty much the next best thing, at a significantly reduced price. 21.5% min eff at 3.1 watts per cell. Our array wattage under max illumination is 55.8w, and we need a bit under 6w to sustain level flight. The key to what we hope will be our success is an very efficient motor prop combination that we've put a lot of thought into.

Look forward to seeing you at DBF again this year ;)

[MIT KID]

So we finished the plane yesterday, and were planning on doing our first flight today. We re-did the prop analysis after we had a more accurate AUW for the plane, and found out that we needed a larger prop to meet our climb requirements. We ordered new props last week but they haven't arrived yet, so we had to fly with our origional prop.

With the origional 16x13.5 we get about 1N of thrust at full throttle which translates to about 70 fpm clim brate. With the new 18x13.5 we get 1.8N of thrust which gives a much less marginal 150fpm climb rate.

Despite not having the proper props we decided to give the flight a go and see how the plane handled. Weather conditions were less than optimal with a sustained 6mph wind with 10-11mph irregular gusts, but we gave it a go.

With me at the controls ( :eek: ), and my partner videotaping Prof. Drela gave the plane a strong hand launch (20+ feet) and I leveled out and applied full power. The plane would hold level flight but was struggling to climb. This was compounded by the irregular wind gusts which were blowing me precariously close to soccer goals and dormitories. I cut the throttle and glided in. We repeated this a couple of times with similar results.

Good News: Plane handled nicely, trim and cg were right on the money. The plane feels very clean, and had an excellent flat glide. WE DIDN'T BREAK ANYTHING.

Bad News: Inadequate thrust with the 16x13 prop. We need to wait for the new props before we can do any more test flights. We also need less gusty conditions for safe flying (which isn't the typical weather in Novermber in New England)

Video: http://www.rcgroups.com/forums/showthread.php?t=590725#post6314603

in the last clip you can see Prof. Drela running after the plane with his anemometer trying to get the flighty speed :)

[tnavressdog]

I'd love to see the video! And of course more video of your upcoming flights with the larger prop too.

[JettPilot]

Are you testing using a battery ? I dont see the solar cells on top of the wing yet :confused:

Looking at the placement of the cells in earlier photos, I see you have a LOT more room to put more cells on if you want to ( Almost twice as many if you really tried). I think you would get much better performance covering as much as possible and having twice the power, even if it does add to the weight :rolleyes:

There is another solar powered sailplane, the SoLong UAV that was constructed with the A300 cells and they were able to cover the entire wing with them. They even cut cells in half to put them in places. See picture below.



JettPilot

[MIT KID]

Yes, initial flight tests are being conducted with a battery, and the aircraft balasted to the weight it would be with the panel onboard. We dont want to risk losing the panel durring the shakedown flights, as we definitely don't have enough time or money to replace it.

Th reason the entire wing isn't covered with cells is money. We had a $500 budget to work with, and sized the array accordingly. Cost was a design variable in our optimization exercise, so the wing has been taylored for both aerodynamic and economic efficiency. Right now there is approximately 25% of the wing coverd with cells, which translates to about 55 watts of power at max illumination. The aircraft needs about 6 watts to maintain level flight. By adding more cells we just bump up the power margin allowing us to fly later in the day, but as is we should be able to fly when the sun is about 6 degrees above the horizon, which is pretty good.

[MIT KID]

So we got the new props in this morning. We re-drilled the balde grip to get the extra pitch we need and headed out to the feild. Climb without the lead balast was good, and we were able to make two flights with multiple circles around the field. We then added the 300g lead balast to simulate the weight of the panel, climb was more marginal, but better than yesterday. There were a couple of feild goal posts in the middle of the field that I was worried about hitting so I set it down rather than trying to make a turn. The battery was starting to sag, so we decided to pack up and head home with a plane ;) .

What this flight showed us is that we still aren't loading up the motor, at full throttle we were only pulling a bit over an amp, and the motors rated for 2 amps continuous. We're having a larger prop overnighted (23x12) we ran it through qprop and the thrust numbers look better. If that doesn't resolve the climb issues we'll look into molding our own prop.

New flight footage is up in the video forum.

[JettPilot]

Very cool MIT KID ! I have always wanted to do a project like you are doing. Build a solar powered sailplane with good performance. The mistake most people have been making is that they want to charge batteries, which more than doubles the power requirements and adds a HUGE amount of weight to the plane.

If you were to completely cover that wing with cells, and use bigger motor, you would have the performance to get that thing to a VERY high altitude :D With electric power, there is not performance loss as altitude goes up, and the energy to the solar panels gets stronger. Ever considered flying that plane to 80,000 feet or so ?

I proposed this last year, and was going to try it, but real life got in the way and I never got to do it :( Check out this thread I made on it last year !

http://www.rcgroups.com/forums/showthread.php?t=370716

I posted some of your pictures and text in that thread, as your project is exactly along the lines of what I was proposing.

Its really great seeing your project and all the pictures, that is just way awesome !

JettPilot

[frankswd]

Too cool for school! I wish we could do this back in my day!

[MIT KID]

So yesterday's test flights were done with an 18.5x13 aeronaut prop with an extra 8 degrees of pitch added in (i'll post some details of the re-pitching process later). But we still aren't loading up the motor, and needed more disk area. The problem is that the largest CRP prop is 18.5 inches, so we finally broke down and ordered one of the carbon freudenthaler props from soaringusa.com, which come in sizes up to 23 inches. We got the 23x12 ($50 plus overnight shipping :mad: ). After entering the pitch distribution and profile into qprop, the numbers look very good.

Last battery powered flight test tomorrow (hopefully) then its on to solar power.

[awmeade]

mit kid - loving your work there, looking very professional! Very lucky too to be work with Dr. D!

I had one small thought as I am unfamiliar with your motor - are the bearings OK do you think to come up to this sort of size prop?

Keep up the good work, looks like you've aced this one :)

[Audax]

MIT KID,

Wow, what a cool project. This may be a moot point if your new prop is big enough, but have you considered a custom machined spinner that would accept 3 or 4 blades?

Have fun,

-Greg

[MIT KID]

Today we flew under battery power with the new prop, and the climb issues seem to be resolved. It was able to sustain an acceptable climb rate ballasted to the solar-ready weight.

When we got back to the shop, I finished wiring up the array (which was one of the most nerve racking solder jobs I've ever done) and attached it to the wing. So, now the aircraft is ready for solar powered flight test, the first of which will be tomorrow around 11am. We need 22 degrees of sun inclination to get our desired climbing power, which translates to flying between 9am and 2pm (11:30 am is max solar inclination) as the sun is quite low in New England these days.

Hopefully I'll have good news for you all tomorrow afternoon ;) .

[JettPilot]

How much did each A300 solar cell cost ? It looks like you were able to curve them, are they flexible ?

JettPilot

[MIT KID]

The bare A300's are $22.50 a piece. But you need to get them encapsulated which adds about 50 to 75% of the bare cell price to the cost. The encapsulation is definitely worth the extra money, it makes the cells much easier to work with and protects them against around the shop wear-and-tear. The encapsulation also distributes loads more evenly over the cell, so it is easier to bend them around gentle curves without breaking.


How much did each A300 solar cell cost ? It looks like you were able to curve them, are they flexible ?

JettPilot

[MIT KID]

We made our first solar powered flights today... Details later tonight

[Tuner]

Cant wait!

[JettPilot]

Cool !! I will stay tuned...

[MIT KID]

Yesterday we attempted our first solar-powered flights. Up to this point we had done all of our flights on Briggs Field (MIT's main athletic field), and we decided to try the first solar flight there. While Briggs is pretty much the largest open area near campus, it is surounded by dormitories that can generate some rather strange air currents. It's normally not a problem, but yesterday immediately after launch we got hit by a large air bubble coming off of one of the dorms. The plane rolled almost 90 degrees to the left, and I had to make a forced landing while trying to dodge backstops and soccer goals :eek: .

We decided that Briggs wasn't the optimal flying environment, particularly when flying on solar power only, so we decided to do subsequent tests at Davis Field, which is Charles River RC Club's glider field in Sudbury, MA, about 35 minutes from campus. This afternoon we drove out to Davis and got ready for attempt number 2 at solar-powered flight. We arrived around 1:30pm and there was almost zero wind and a cloudless sky. We launched flying away from the sun and banked left to get the array pointed at the sun. Our power point tracker isn't finished yet, so I spent a few laps flying in ground effect, trying to find the motor's sweet spot. After I got the throttle figured out, the plane started to climb. Power was at its maximum flying away from the sun and turning towards the sun, so most of the climbing was done during these portions of the circut. Turns away from the sun were kept very flat to keep some light on the array, but it was nearly impossible to climb in this situation. It would be nice to have a more open area to work out flying strategies, but Davis is sufficient for now.

We made two flights lasting about 8 minutes each at 1:45 and 2:10 (solar inclination angle of about 20 degrees). Durring both flights we were able to achieve a decent climb rate, which would be considerably stronger earlier in the day. Based on today's results, we should be able to sustain climbing flight from 9 am to 2:30 pm, which is pretty good for November in New England. Too bad we weren't testing in the summer months when there's a bit more daylight.

All in all my partner and I were very pleased with today's flights, and look forward to pushing the limits of this design. Thank you all for your support. I'll post updates as we move forward with flight testing.

Videos:http://www.rcgroups.com/forums/showthread.php?t=593296#post6346026

[planebane]

Congratulations on your successful flight. It must feel great! This is a fascinating project.

[rroback]

is drela an advisor? That tail looks quite familiar ;)

Rhett

[Mihai]

Cool project indeed! A couple of questions/comments:

1) Did you consider putting three cells on the wing tips? This way you'll likely reduce the max power, but will have more at lower sun angles. In the same idea you can put some on the vertical tail (rudder?) - you'll get very low sun angles working there.

2) How do you plan to test its endurance? By RC? This is unlikely to be optimal: if for nothing else the range sucks - you can't go all the way "up". You'll speck out of control. Furthermore, it's likely that the best path is long climbs in the optimal direction with short turns. I mean LONG climbs (several km).

3) Finally, you realize that it's very unlikely to break any solar records in the winter in MA :-) I'd try the north pole in the summer :-). If you time it right you may be able to send it there with a properly programmed autopilot and have it return in 6 months. How is that for a record?

Really nice work. I assume that you're in the aero- department, and hence, quite far from an autopilot, right?

All the best,
M.

[MIT KID]

Thanks for the comments everybody.

Rhett- Drela is our faculty advisor for this project, and yes the tail feathers are from a from an EZ bubble dancer kit from polecat aero. There is considerable lightening and thinning of their profiles. We were able to cut almost 15g out of the tail group compared to a standard bubble dancer.

Mihai- 1.) The cells were kept towards the center of the wing to keep the roll inertia low. It is only rudder/elevator control, so if we put them out on the tips the roll authority would greatly suffer. Also given our long tail boom and the considerable weight/size of the cells, mounting them on the rudder was pretty much out of the question. It would have been nice to have more cells, but again we were constrained by cost.

2.) Endurance numbers (at least for now) will be based on our powere levels at different sun angles. If our data shows that our power drops below cruise power at say 3:14 pm we can correspond that time to the suns position for that day. This will allow us to predict the times of day that we can opperate the aircraft and give us a projected endurance for any given day of the year. For record attempts we might consider using something in the 2.4 Ghz range (i think the DXZ claims a 5mile range)

3.) Yes we are aware that there will be no record attempts until the spring/summer. I don't even know if there is 9 hrs of daylight in MA these days. We need to collect data now, however, because this is part of our senior design class required for graduation :eek: (groups of 2 student develop and execute an experimental project). We had a similar tought about the north pole thing, we'll probably just drive up to canada sometime around the solstice.

Yes, I am in the aero dept, but not so far from autopilots. I work as an undergrad researcher in the Aerospace Controls Lab, and we have quite a few uav autopilots. The problem would be trying to shoehorn something into out tiny fuselage.

-Adam

[Mihai]

What's the top speed on that thing? How far up can you theoretically get (given good sun)? And how "flat" can you glide (a rate of descent while gliding). I'm trying to figure out how far are you from going forever (storing potential energy rather than in batteries). I'm also trying to figure out how close to the pole you need to go before it can fly "all summer". Can you post (send me) the report once it's done? I'm really interested in the details.

Regarding the placement - forget it. Somehow I thought that the cells are really light. I played with some cells in my lab and they were quite light (not your brand though), but your plane may also be *very* light and get messed up even by a light cell. To make a long story short - how heavy is the plane and how heavy is a cell?

Finally, it seems that you make the life hard on yourself by not allowing to use batteries. Was "forever" flight demonstrated while using batteries?

M.

[Taylor Mullin]

Instead of getting a huge prop, why couldn't you just get a new gearbox with a smaller gear ratio?

[Carl.E]

First let me introduce myself. I'm Carl, Adam's partner on this project. I'm also a senior studying Aero/Astro at MIT. I've been building and flying R/C since I was 11. I finally got around to registering on RCG, hence this being my first post. I've been mainly been on that 'other' R/C forum until now. Ok, now on to some question answering...

Mihai,
The motor/gearbox/prop combo we have on there is optimized for max efficiency at the minimum flight power speed range of 5.5-6 m/s. We've been mainly focusing on that aspect, so we don't have the calculated top speed off hand; but it's probably around 8 m/s. The high pitched, slow turning prop gets out run by forward speed pretty quickly.

Same goes for climb rate and maximum altitude; in fact, our maximum power output of the motor is only about 20 watts at full sunlight! We have the (55 watt) array way oversized specifically for flying at low sun angles. Motor efficiency starts to suffer for any given motor that isn't loaded at least some non-zero fraction of it's design power; if we had a 55 watt motor installed, it wouldn't be at its max efficiency at 8 watts where we need peak efficiency to maintain flight. This all means that once the sun gets over about 35 degrees or so, our power output stops increasing. The increased efficiency at lower power levles allows us to fly on that much lower sunlight levels. However, given that our plane has been designed for maximum power margin, a simple motor/gearbox/prop swap to something that can take advantage of the full array power output efficiently is all that is needed to make this into a climbing and (moderately) high-altitude machine. Probably a small lightweight brushless. Losing 4 ounces (10% of our total weight!) will also improve climb performance. :rolleyes:
As for glide performance, our L/D is little over 20.

The plane ready to go minus the array weighs 930 grams. The array (18 encapsulated cells covering about 1/3 sq. meter) weighs 350 grams. The cells are really light; it's just that our plane is too. ;)

"Forever flight" was demonstrated by SoLong last summer during their 48 hour flight using 225 watts of solar cells and 5.6 kg of Li-Ion cells. There are a couple links to threads about it up near the beginning of our thread. Our eventual goal with this plane is to break the FAI model aircraft endurance record (and maybe others) for solar-powered flight, which states that there can be no power buffer (read batteries) between the array and motor.


Taylor,

We'd take an efficiency hit by switching gearboxes to spin the same prop faster to get more thrust. At the low speeds and thrust levels that we are at, props are more efficient if they are large and turn slow than if they are smaller with higher RPM. It's an energy vs. momentum thing. Thrust = mass*velocity, but energy = 1/2*mass*velocity^2. So to get a given amount of thrust, it is more energy efficient to move a big chunck of air a little bit than to move little bit of air a lot. Take a look at the prop on any of the human-powered aircraft or other minimum-power designs; it's big and slow-turning. Of course there's a practical limit to this in that the off-design performance suffers...


...Wow, that was one heck of a first post!

-Carl

[Mihai]

Carl, that was a heck of a first post! Thanks a lot for clarifying a few aspects. One remaining question is concerning the possibility to use an "altitude buffer", i.e., instead of storing energy in batteries over night to store it in potential energy by climbing as much as you can and then gliding as straight as you can (toward east! :-) until the sun comes out. To figure out if this even in the ball-park, I need to know your best glide ratio and maximum altitude. If it's something like 10:1, at 5m/s then you need to get it up at least 18 km to cover something like 10 hours of night. If it's 30:1 you only need 6km. Is this possible?

Thanks,
Mihai

[philthyy]

Man I love this forum! This is such a cool project. I never knew that solar cells had advanced enough for this kind of project to even be possible at a college level. I knew the military could do stuff like this with a budget we cant even begin to imagine, but its really cool to see solar cells cheap enough for hobby use! Hmm...gonna have to start planning my own project....

[dfratello]

Great work - congratulations on your success!

The bare A300's are $22.50 a piece. But you need to get them encapsulated which adds about 50 to 75% of the bare cell price to the cost. The encapsulation is definitely worth the extra money, it makes the cells much easier to work with and protects them against around the shop wear-and-tear. The encapsulation also distributes loads more evenly over the cell, so it is easier to bend them around gentle curves without breaking.
I'm an engineer advising a jr. college team working on a terrestial solar application. Can you tell us where you got your encapsulated A300 cells -- from Sunpower or a third party?
Appreciate any info you can provide.

Dave

[MIT KID]

just pm'd you our supplier's contact info.

Great work - congratulations on your success!


I'm an engineer advising a jr. college team working on a terrestial solar application. Can you tell us where you got your encapsulated A300 cells -- from Sunpower or a third party?
Appreciate any info you can provide.

Dave

[parky]

Carl, that was a heck of a first post! Thanks a lot for clarifying a few aspects. One remaining question is concerning the possibility to use an "altitude buffer", i.e., instead of storing energy in batteries over night to store it in potential energy by climbing as much as you can and then gliding as straight as you can (toward east! :-) until the sun comes out. To figure out if this even in the ball-park, I need to know your best glide ratio and maximum altitude. If it's something like 10:1, at 5m/s then you need to get it up at least 18 km to cover something like 10 hours of night. If it's 30:1 you only need 6km. Is this possible?

Thanks,
Mihai

In general, batteries are the better option, particularly if you want a practical airplane that can operate for most of the year. Getting enough altitude to stay aloft whle gliding through a 14 to 16 hour night is a major problem. And when you do, you get very high airspeeds at high altitude, and that gives a high power requirement and high sink rate. The battery (or fuel cell) system is heavy, but you can fly constant altitude and with a low minimum power all night long. FWIW, I was a co-author on a NASA study that looked at all this, published in 1982! (CR-3699). We were mainly looking at high altitude (20km), all weather solar airplanes, but I think the same trends will apply to low altitude as well.

What is interesting is that, as long as you can do the right power conditioning, LiPoly batteries have really awesome energy storage efficiency (i.e. discharge power out/charge power in).

Bob

[Mihai]

Sure, I have no doubt that batteries are a better option, but they want to go the "righteous" way and not use any. Then I don't see any other way to store energy but to gain altitude. I would assume that since perpetual flight with batteries was only recently demonstrated and it's much "easier", just gliding though the night is not (yet?) possible. I just wanted to figure out how far out it is.

Best,
Mihai

[dalbert02]

A year has gone by. Anything new? How did your project end? Was your hypothesis proven? Were there any unexpected results?
-dave

[dmgoedde]

A year has gone by. Anything new? How did your project end? Was your hypothesis proven? Were there any unexpected results?
-dave

dalbert02 - I was going to offer these guys a complimentary AttoPilot to use in their plane when I read that none of the autopilots they considered were small enough to fit in the fuselage. I read through the long post, and determined that these guys graduated last year.

I do hope they see your question and post a reply. Such a neat project.

[Carl.E]

Wow, I was just browsing through RCG and glanced in the UAV forum and saw our solar thread was recently replied to! I'll take the enduring support as a complement. :) Be warned, this is going to be a long update.

Yes, a year has gone by, and a lot has happened. We finished this project very successfully that semester (even won a department award for it! :D ), lead the MIT AIAA Design/Build/Fly team to victory in the spring (short thread about that here: http://www.rcgroups.com/forums/showthread.php?t=677315 ), and then both graduated in June. Adam (MIT KID) started work at Aurora Flight Sciences in Cambridge, MA working on some really cool conceptual designs and proposals. I took a job with Boeing Advanced Unmanned Systems - Concept Exploration in Irvine, CA, working on the A160T Hummingbird, a full-scale (36-foot rotor diameter) UAV helicopter.

Shortly after graduating, we made an attempt for the FAI Solar Endurance record. Due to schedules and logistics of getting official observers and an unused field (especially tough on a perfect flying day!), we had limited days open to try. June 19th looked to be our best shot weather-wise, so we aimed for that day. Long story short, the early-morning winds here higher than expected, and while we were able to sustain flight for short periods, we were unable to climb through low-altitude turbulence before our determined cut-off time for breaking the record. It was a good learning experience for any future FAI attempts.

Soon after the record attempt, I moved back home to Wisconsin for a couple months before heading out to CA to start my job. Adam and I decided that Southern California has much better weather for a solar airplane than New England, so I adopted the plane and brought it with me in case we decide to try for record flights again.

In early November, I got a chance to fly the solar plane out here in CA. There is a beautiful field in San Pedro called the Field of Dreams that is reserved for electric flying a couple Sunday's a month. One of our department's senior advisors has actually taken a job out here, and invited me to come fly at the Field of Dreams. The solar plane flew once again, just shy of one year from its first solar-powered flight! I flew it for ~20 minutes; all the other guys at the field loved it and were quite impressed. It was a very strong flight and was able to climb pretty easily. During one pass, I was hit by a radio glitch, which commanded full down elevator briefly. I was able to save it, but caught the prop on the grass as I was pulling up. It cracked the motor mount a little, but otherwise was undamaged. It was a great day overall, and was exciting seeing the plane back in the air. It encouraged me to develop a 2nd propulsion setup that I had been thinking about for a while.

I'm currently working on a small brushless setup that is able to use the full power available from the solar array. The current brushed motor system is designed for maximum efficiency at cruise power (8.5 watts to the motor from our flight data...predicted was 8.2 watts). The drawback of this setup is that it can only sustain about 20 watts at full throttle (motor limited)... not much excess power to climb in turbulence, as we found out! We took that hit to squeeze every last percent of efficiency at low poweres from the propulsion setup for our design. This was done to maximize flight time by allowing flight at lower sun angles. I'm now sizing a propulsion setup that emphasizes flyablilty and better climb performance at the expense of a few percent efficiency at the lowest power settings (ie super low sun angles). This one will be able to sustain the full 56 watts from the array with ample sunlight. It will make the plane much more of a cruiser and less stessfull to fly. It will most likely not be as efficient during low-power cruise or be able to fly at as low of sun angles. However, it will allow it to take advantage of any time the array is able to put out more than 20 watts. Another plus is that the plane will lose 3-4 ounces of motor weight, which will also aid in climb rate. I'm designing this 2nd system to be fully compatible with the current motor mounts so I can swap it for the original setup easily.

I'll do my best to keep updating this thread as I select this motor/prop combo, build it, and test it. I'll definitely post with any more flights and how the new motor works out!

dmgoedde, I really appreciate your support! A small simple autopilot couple be used to our advantage for such record attempts as maximum altitude as well as endurance and distance. Could you get me more info on the AttoPilot?

-Carl

[rroback]

Sounds pretty awesome. You guys definitely did well at AIAA DBF last year, and we certainly had our difficulties! Irvine is more pleasant then Cambridge, but it's a lot less exciting!

[dalbert02]

Wow! Great work! Keep the posts up! I wish I would have gone to MIT, sadly they rejected my application... :(
-dave

[Jack Crossfire]

Now U know who the Lockheeds & the Boeings R hiring to do this stuff for a living.

Was at a company which hired an MS in CS from MIT. They were losing money and getting rid of people constantly. She got promoted to upper management in 2 years. Those R really smart cookies.

[CAFplanekid]

I was thermalling this summer a while back and this thread popped into my head, thanks for the update.

rroback: I never got to see one of your flight attempts (always repairing our plane in the tent, haha), but heard you guys had some problems. Ever get that figured out? I know all too well what a PITA that configuration was. Are you in it again this year?

Jonathan (Purdue)

[rroback]

Well, we got our issues basicall figured out, but it was too late. flight one was short, as a certain person didn't instal the receiver pack correctly, on an extremely cg sensitive plane, so we did a loop and crashed. spent all day and night rebuilding ( till about 4 am), and flew the next day, and got most of the way around the course before all hell broke loose. I couldn't see the plane, but interference seemed like the end result. Our super light plane, with a 23inch span...boy.. that wasn't fun to fly. I'm doing it this year.

[CAFplanekid]

Ah, I've lost a plane due to a receiver pack coming loose in flight, always the little things.

I never flew our BWB, but know it wasn't any fun to fly. This year hopefully won't be so bad. I'll be the pilot for this year.

[Carl.E]

Time for an update:

I've selected a brushless setup for max climb rate while maintaining a high cruise efficiency as I mentioned in my last post here. I analyzed a number of motor/gearbox/prop combos and looked at their power vs. airspeed curves as well as their efficiencies at min cruise power using QPROP.

I ended up choosing the Mega 16/7/17 with a Kontronik 5.2:1 gearbox (awesome quality!!) turning the modified 23x12 carbon folding prop we used on the original setup. That prop is 79% efficienct at cruise, and actually increases to 82.5% in the optimum climb conditions! Max climb rate is projected to be 410 ft/min at 8.5 m/s, compared to the brushed setup's ~90 ft/min. Efficiency at min cruise power is slightly lower than the original setup, but the power savings from having a lighter (3 ounce savings) plane will help offset the efficiency hit.

I have received the motor and gearbox, but ordered the wrong gearbox adapter, so I can't assemble it until the new one arrives next week. :( I'll start work on modifying the mounts this weekend, since the motor and gearbox diameters are a couple mm larger than the Maxon setup. I should have the plane ready to fly with the new setup by the end of January. I'll have more updates I'm sure. :)

-Carl

[Aio_1]

...Max climb rate is projected to be 410 ft/min at 8.5 m/s, compared to the brushed setup's ~90 ft/min...
Ahhhhh:eek:
Mixed systems of measurement in one sentence and both for speed. It's sacrilege! ;)

Aidan

[Carl.E]

I admit, that's the product of an American engineer that went to a college that teaches mostly everything in SI units... Because of that I convert units to what I'm comfortable and familiar with all the time... I know how fast 8.5m/s is (pretty slow) and I know how good of a climb rate 400ft/min is for a glider. But tell me the speeds with swapped units, 1670 ft/min velocity and 2.1 m/s climb rate, and I couldn't quickly say if those were reasonable at all! :rolleyes:

[macboffin]

Ahhhhh:eek:
Mixed systems of measurement in one sentence and both for speed. It's sacrilege! ;)

Aidan How many leagues a day is that again?

[Capt. Crunch]

8.5 m/s = 51,109.418 928 furlong/fortnight = 152.111 362 league/day, or 19.014 mph

But my question is what is your stall speed in kph or Knots? Methinks you're too close! Ouch...

Just read back on the thread and see your stall is below 5 m/s, good, strike the above comment.

BTW, I believe for FAI & AMA records you can't use an auto pilot, and must keep the aircraft in sight at all times. Does Wolfgang Schaeper still have the pure solar altitude record? Also duration I believe. Of course he cheats and uses cells from his space project arrays, blems, of course!

I think it is just awsome that you've done this on $500, that alone is a record of some kind. I believe I can be an FAI Official Observer, and I'm sure Steve Neu is, so there would be two OOs if you want them for record attemps this summer.

Best of luck and keep this thread posted and it'll send updates to those on it when something new is added.

Wayne

[Carl.E]

Does Wolfgang Schaeper still have the pure solar altitude record? Also duration I believe.

He does have those two records still. Also distance in a straight line, speed in a straight line, distance in a closed course, and speed in a closed course. He currently holds every FAI F5-SOL record there is: http://records.fai.org/models/current.asp?id=199

Thanks for the kind words about the project Capt. Crunch, as well as volunteering to be an official observer for any future record attempts! :-) I'll definitely keep you (and this thread) posted.

On that note, I got my motor and gearbox assembled, and I just finished modifying the motor mount rings to accept the new motor this evening. The firewall epoxy is curing as I type this. I should have the motor installed and ready to go tomorrow, assuming I have some time.

I'll post some pics when I can...my camera seems to be broken...then lens mechanism is jammed. :-(

-Carl

[Magician]

Unless the FAI has updated their rules recently, you can use an autopilot and not have eyes on at all times. I'm no expert but Maynard Hill's record flight from Nova Scotia to Ireland used an autopilot and I know he didn't see it the whole way :) That flight holds the longest duration and longest distance for piston powered model airplanes.

It's nice to see a DBF alumni out in the real world and earning a living in the UAV industry. Maybe they should ask for your help to keep the A160 in the air!

Regards,

Chris
(DBF flight line fixture)

[radiohound]

Unless the FAI has updated their rules recently, you can use an autopilot and not have eyes on at all times. I'm no expert but Maynard Hill's record flight from Nova Scotia to Ireland used an autopilot and I know he didn't see it the whole way :)

Now that is funny!

Even if his eyes were on it, he is legally blind, so it didn't give him much advantage. You probably already knew that, but others may not. Pretty impressive!

[Carl.E]

The motor is installed and ready to go. All that is needed before flying is to check the balance and some nice sunny weather (it's been rainy and windy here in SoCal for the past several days). Balancing the plane will require everything to be pushed as far forward is possible, plus a little nose ballast maybe, since the new motor is 3 ounces lighter than the Maxon setup. I hooked up a 2-cell Lipo to run up the motor, and boy does it pull hard! Way more thrust at 7.4 volts than the original setup had at even 11.1 volts! Can't wait to get this thing in the air!! I've attached a couple pics showing the motor size and installation differences.

-Carl

[Carl.E]

Got the plane flying with the new motor setup yesterday! All I can say is it's awesome! :D Instead of needing every spare watt to fight for every foot of altitude I can get, the new system climbs up with authority. I could climb at will--a first with this airplane! Still able to throttle back and just cruise around using very little power as before, which will be essential during low sun angle flight for the endurance attempts. We had an Mini EagleTree data logger installed, so I've got power and RPM data to look at too. I ended up doing 2 flights; one was 22 minutes, and the second was 31 minutes. They could have easily been hours, but the recorder memory only had ~15 minutes of capacity, so I decided to bring it back down to upload the data.

Everything looks promising for some record attempts this summer. Next set of flights will probably be to tweak the cg, explore launches at low sun angles, and do extended flights of a couple hours. I didn't get any video yesterday, but a couple of pictures are attached below.

-Carl

[rroback]

Great work! Where were you flying in Irvine?

Rhett

[Carl.E]

I wasn't in Irvine...you think there are any fields in Irvine that are that big and open with nobody there on such a nice day as Saturday was? :p

The field I was at is actually the drill field at Fort MacArthur, up in San Pedro. Pete Young, who is a retired airforce Colonel as well as MIT Aero/Astro lecturer, has recently moved out here and has been flying with me. He has access to the base where this jewel of a field is.

[Capt. Crunch]

Looks great Carl, lucky on the field too!

looking forward to more flights.

[gtfreeflyer]

Carl, have you checked out Fairview Park in Costa Mesa? Lots of flying space there.

[lobbwill]

wow this is an amazing looking project - what output are you looking at from each solar cell? Im looking at some Unisolar super-light ones which are rated at 1.5V and 300mA - that's not even half a Watt tho?!

Will

[Carl.E]

gtfreeflyer,

I have been to Fairview park in Costa Mesa--a very nice field, and nice people too. I haven't had a chance to get a permit from the City Hall, which is required to fly there now, however. Also, the 400 foot ceiling is a little restricting. I do plan on getting a permit and perhaps flying there. I do know of another solar airplane that flies there, though.


lobbwill,
Thanks for the kind words; it has been quite an exciting project. Each solar cell puts out ~0.67 volts open circuit, and 6 amps short circuit current. At the max power point, each cell is at about 0.55 volts, and just under 6 amps, so about 3 watts a piece. There are 18 strung together in series in the array, giving roughly 55 watts of power, operating at just over 10 volts.



I flew the plane last weekend again, and had some interesting results; I'll post more about it later (probably this evening) when I have some more time.

-Carl

[Krogoth]

Just while the topic's active -

I happen to be a final year engineering student, and I'm looking at combining ideas from your project with the hyperblimp concept, to create an 'aerosatellite' - one that remains in the air 24/7, relaying data and live video from around 1500ft, and survives off solar cells and LiPoly batteries.

Carl, I just wanted to know a couple of things:

1. I found in the forums that your solar cells generate 3W and cost 22.50, but how much does each weigh? I'm trying to get the optimum weight-power-cost ratio, to calculate the size of my blimp. Have you got a datasheet?

2. What kind of efficiencies and thrusts were you getting from your brushless configuration? My plan was to design the blimp to operate continuously at 10m/s (to oppose wind etc), do you happen to have any figures of thrust per watt for your motor/prop configuration?

Again, congratulations on a fantastic project, and I look forward to hearing from you.

Will.

[Carl.E]

Krogoth:

1. The total encapsulated array weighs about 350 grams. Divided among 18 cells gives about 19 grams per cell. Remember this includes the encapsulation material, which isn't light. I don't have an actual number for you, but I'd guess 10-12 grams per bare cell. However, getting your hands on bare cells is nearly impossible, so you will probably have to get them encapsulated by a 3rd party.

2. I haven't measured efficiencies and thrust for the brushless setup like we did for the brushed setup, since I no longer have access to a wind tunnel and thrust stand. I have the theoretical numbers if you want, just let me know.

Also, thrust per watt isn't really an engineering efficiency, since it doesn't take into account airspeed. Static thrust is what most people use for thrust per watt, and how often do you fly at zero airspeed? (Not including hovering! :p ) What you really care about is thrust*velocity per watt, which is what efficiency really is: Power_out / Power_in.

-Carl

[Carl.E]

Flight update:

Last Sunday, I went and flew at Field of Dreams at San Pedro again. The first flight was launched with the sun behind some hazy thin clouds, and the lack of power was very noticible. However, I was able to stay up and fly several laps at low altitude. The data showed an average power of ~9 watts to sustain flight...a little worse than the brushed setup, but not bad.

Once the sky cleared, we launched again, and the plane climbed beautifully. After ~20 minutes into a solid flight, the motor seemed to lose power and the airplane slowly lost altitude. After landing, I realized the ESC had gotten really hot and desoldered one of the motor leads! :eek: I was only running ~2.5 amps through the Phoenix-10. I resoldered the lead and did a quick ground test, and sure enough, the motor got too hot to touch after 5 minutes at partial throttle! A little RCG research that night found that ESC's are actually pulsing full current through the motor during partial throttle. Full throttle current with my setup is still only 6 amps, so I was unsure of the problem. I posted a question in the Castle Creations about what happened here: http://www.rcgroups.com/forums/showthread.php?t=820724
Turns out the PHX-10 doesn't play nice with very high inductance (high wind) motors due to it's slower switching high-side FETs. Patrick suggested a PHX-25, since it's more suited to high inductance motors. I have one to try but haven't had a chance to install and test it out yet. Hopefully a cooler running ESC will produce lower required flight powers!

-Carl

[Carl.E]

Got another chance to fly this past Sunday. Beautiful weather: clear skies and very low winds all morning. I had installed and tested a Pheonix-25 as suggested in the Castle Creations forum after desoldering the Pheonix-10 I was originally using. The tests went well; minimal ESC heating at partial throttle, hardly warm to the touch after 15 minutes of running.

First flight we launched at 10am with a very nice climb-out...one of these times we'll get it on video, I swear! In addition to the EagleTree Micro E-logger, we installed an altitude logger. This flight was to verify satisfactory ESC temps, so I kept it short at about 15 minutes. After landing I checked the ESC, and it was hardly warm, just like the initial testing showed. Success!! :) However, we failed to properly set up the temp probe on the data logger, so no actual temps were recorded. We did get some very nice altitude data, though. After the flight, I was able to overlay the altitude data with the power plot, which showed a nice correlation. It's attached below. Max altitude was ~300 feet.

The second flight was planned to be an hour duration--the longest flight to date. After launching at around 10:45, I settled back into a lawn chair to get used to flying sitting down... a strange feeling at first for those of you who have never done it before! Everything was solid; the plane handled beautifully and climbed at will. An hour came and went, and I felt great, so I decided to push on for 90 minutes or more. As you can see from the altitude plot below, I was able to climb pretty high on this flight....slightly over 1,200 feet! A 10' span wing sure looks small that high up! You'll also see at one point the plane suddenly dropped from 1,000 feet to 400 feet in a matter of seconds. The steepest descent rate was about 3,500 feet per minute! :eek: I'm a little unsure of what happened... the plane seemed to drop nose down some and just fall....maybe a large pocket of sinking air, radio lock-out, or just pilot error letting the nose fall too much without correction? Luckily I was able to come out of the descent and continue my flight. So much for flying for extended periods being boring! After things calmed from this, I decided not to push my luck too much more, so I landed after a flight time of 2 hours, 2 minutes.

We went to retrieve the flight data for the 2-hour flight, and found the logger was empty! No power or temp data to be had for this flight. :( Also, the altitude logger memory filled up about 80 minutes into the flight and stopped. I'm pretty sure I had a 3rd climb up to ~1,000 feet in the last 1/2 hour of the flight that was not recorded. The motor and ESC were warmer to the touch than after the first flight, but not too hot to hold my finger on them. This was the first time the motor had gotten warm at all. I may have to cut some small cooling holes to get some airflow over the motor. The spinner has a hole in the front for cooling that will work perfectly if I can carve away some of the firewall where the motor mounts up.... I'll look into it.

Lessons learned for future endurance flights:
Remember to bring sun screen.
Sit in the shade if possible.
Don't wear a black shirt in the sun for 2 hours.
(Yes, these above observations lead to me getting pretty sunburned. :rolleyes: )
Double-check that loggers are working and are sampling slow enough to cover the entire planned flight.

Next flights:
Along with more time-building flights, I'm looking into possibly doing a cross-country flight record attempt as a stepping stone to an endurance record attempt in June. The current FAI straight-line distance record for a solar R/C model is only 30 miles. Easily covered in 2 hours. Mostly a logistics and planning issue it seems. I'll be doing some more research into this.

Any questions/comments/suggestions are welcome and appreciated!

-Carl

[spirko]

Carl,
Great work with this aircraft, two questions if you can help:
1) where did you source the a-300 cells?
2) is there a layman’s formula for where / what capacity diodes to use so as to not ruin the solar cells?
Good luck with the record attempt next month, I’m in Los Angeles if I can be of any assistance.
Thanks,
Craig

[Carl.E]

Craig,

For now, PM or email me and I'll get you my contact for the A300 cells. I emailed him a few days ago asking permission to publish his info on here but he has not responded. When I hear back, I'll post it if I can.

As for the diodes, I am unfamiliar with how they are incorporated into the array. When our array was made, they were included if they were needed. Remember that we are not running a propulsion battery, so there is no current available to go back through the array if they stop putting out power.

Thanks for the offer for support. I may be asking for some observers for the attempt in the near future.


I'll be posting an update with my progress since my last post when I get some time.

-Carl


Carl,
Great work with this aircraft, two questions if you can help:
1) where did you source the a-300 cells?
2) is there a layman’s formula for where / what capacity diodes to use so as to not ruin the solar cells?
Good luck with the record attempt next month, I’m in Los Angeles if I can be of any assistance.
Thanks,
Craig

[Carl.E]

I haven't had much time to keep this updated lately, so here's a quick overview of progress.

Last weekend in Perris, CA, I flew for 3 hours straight with no issues. Launched at 8:35am and the plane easily climbed. Got some great data from the data logger, and it looks like the power point tracker is doing its job! Hopefully I'll get some time to post some data. Also, temperatures are under control and the cooling holes I made seem to be very effective even in 100 degree heat; right after landing the ESC and motor temps climbed 20-30 degrees, even with the motor off! Tomorrow I'll be going out and attempting a launch at 6:45, which is the predicted minimum sun angle needed to sustain flight. Adam (design partner and starter of this thread almost 2 years ago) will be flying out to CA and we'll be making an official endurance attempt at the end of next week or next weekend. Hopefully the weather cooperates for us to make an attempt while he is here!

On that note, we are in need of a couple of AMA observers for the attempt to make it official. No special status required, just be an AMA member; we already have our Contest Director who serves as the "official" observer; the FAI rules just require two more people. You would have to be present from launch (6:30-6:45am) through landing (7-7:30pm). However, you can make a flying day out of it; we don't need you to keep your eye on the plane every second of the day. :rolleyes: Anybody interested and available should PM me for more info. The window we are looking at is Thursday, June 26th through Monday June 30th. Even if you are only available one of those days, that would be great! You'd be a part of making history. :)


Also, here is the contact info for our solar array supplier, now that he has given permission for it to be posted. One note that he wants to be made clear: he is very busy, and it's a small operation, so there can be a serveral month lead-time on ordering an array. With that, here is his info:

SunCat Solar is the name of the comany, and our contact in Alain Chuzel (pronounced "Allen"). His email address is ahmchuzel<at>aol<dot>com. (replace the <at> with @ and <dot> with . ) He can fabricate pretty much any size or confguration of array, given a good drawing (and enough money ;) )


-Carl

[Jeffery]

Glad to hear you guys are making progress, Carl. Good luck on the record attempt, if I was closer I'd be a spotter for you.

[wish]

Hi Carl,

I've been reading this thread whit much please! I'm also planning to build a simular PV powered plane... one day...

How did the record attempt go?

Regards from Germany,

Bob Hopman