Help! Seeking prop/motor help breaking the world endurance record for UAVs <50kg with a prandtl wing

[Keaton Hare]

Hi all,
I'm building a solar powered perpetual endurance UAV, ideally to break the world endurance record for sub 50kg airframes. I have been playing around with ecalc and it seems as though a single-engine, 3C perhaps geared design is the best. configuration will be a flying wing with a dual bladed pusher prop extended well behind the wing. Wingspan is 13 ft with a tentative weight of about 15lbs. So far, a 400KV motor with a big prop and maybe a gearbox seems the best option, but plugging a large pitch prop into ecalc gives me a warning about blade stall. Any advice as to powertrain selection and where to find a good reliable gearbox (preferably 5:1 planetary if possible). Will be using a Prandtl wing so no need for a tail or differential thrust. The design speed is very low (14 mph) and just slightly above stall for best efficiency. Any advice from those of you who have experience making efficient designs would be very welcome.

Thanks!

 

[HVB79]

If I were designing it I would try to avoid using a gearbox if at all possible. The extra weight, gearbox losses and reliability problems could cancel out the advantages of a larger prop. Your power system is going to need to run non-stop for a long time so reliability is very important.

How many days is the current world endurance record for a solar powered sub 50kg airframe?

 

[Air-headed Aviator]

If I were designing it I would try to avoid using a gearbox if at all possible. The extra weight, gearbox losses and reliability problems could cancel out the advantages of a larger prop. Your power system is going to need to run non-stop for a long time so reliability is very important.

How many days is the current world endurance record for a solar powered sub 50kg airframe?

The world record according to google is 81 hours! and held by these people http://www.kostasalexis.com/solar-p...Continuous Flight,for any aircraft below 50kg.

 

[HVB79]

The world record according to google is 81 hours!

Interesting, thanks for the link.

Not as long as I was thinking. Some of the larger high altitude systems have 25+ days of continuous flight. Still 81 hours is a long time. Looks like the limiting factor for low altitude systems has more to do with the weather than the available power. The high altitude systems can just fly over all the bad weather.

After looking at their website & YouTube I think they used a nose mounted direct drive motor.

 

[quorneng]

*1 on avoiding a gearbox even a planetary one. Remember at some point the torque required to drive a big prop has a significant effect on the efficiency of the airframe to counter it.
Prop stall should only be a problem on take off if the maximum efficiency flying speed requires a paridularly coarse pitch prop. At 14 mph I would not expect it would.
The power to fly is directly proporional to weight so the lighter the better.
Are you sure a tailess design will give you the best practical airframe? At the size and weight you anticipate there is a danger that to achieve reliable auto stability you may find the efficiency pf a flying wing has to be compromised.
The most efficient flying machines yet produced (high perforamance gliders) all use a conventional tail configuration. Such gliders concentrate on aerodynamic efficiency.

My own modest "endurance" developement was to get a 48" (1.2m) span plane carrying a 5Ah 2s LiPo to maintain height using an average of just 7.5W (1A). Apart from the flight motor that included the power to drive three servos, the speed controller and the radio.
It took nearly two years of experimentation and rebuilds to achieve. The end result was a fairly conventional layout, very light, but with much effort placed on limiting induced and parasitic drag whilst maintaining adequate stability and controlability.

Your project will be many times bigger than mine. You will need persistance and some luck!
Do keep us informed on how you get on..

 

[Keaton Hare]

*1 on avoiding a gearbox even a planetary one. Remember at some point the torque required to drive a big prop has a significant effect on the efficiency of the airframe to counter it.
Prop stall should only be a problem on take off if the maximum efficiency flying speed requires a paridularly coarse pitch prop. At 14 mph I would not expect it would.
The power to fly is directly proporional to weight so the lighter the better.
Are you sure a tailess design will give you the best practical airframe? At the size and weight you anticipate there is a danger that to achieve reliable auto stability you may find the efficiency pf a flying wing has to be compromised.
The most efficient flying machines yet produced (high perforamance gliders) all use a conventional tail configuration. Such gliders concentrate on aerodynamic efficiency.

My own modest "endurance" developement was to get a 48" (1.2m) span plane carrying a 5Ah 2s LiPo to maintain height using an average of just 7.5W (1A). Apart from the flight motor that included the power to drive three servos, the speed controller and the radio.
It took nearly two years of experimentation and rebuilds to achieve. The end result was a fairly conventional layout, very light, but with much effort placed on limiting induced and parasitic drag whilst maintaining adequate stability and controlability.

Your project will be many times bigger than mine. You will need persistance and some luck!
Do keep us informed on how you get on..

Hi Quorneng

I suggested a gearbox because the Atlantik solar team used one for the current 81hr record. Here's a paper detailing their design process. They used an obscure german 400kv motor, 5:1 gearbox and a .66mx.6m prop custom designed. We are working with the former head of R&D at hartzell and Al Bowers, who designed our prandtl wing. The prandtl wing has about 20% greater efficiency VS a traditional wing and produces proverse yaw, so as a flying wing it is inherently more stable than a traditionally configured airframe. The other benefit of a flying wing is that there is 0 parasitic drag which makes design quite simple. I think my specific concerns were about a prop with that severe of a ratio and its advantages compared to one with a shallower pitch. I would absolutely love to hear more about your airframe! do you have a write-up or some specifics posted? That powertrain in particular sounds super interesting

 

[Keaton Hare]

Interesting, thanks for the link.

Not as long as I was thinking. Some of the larger high altitude systems have 25+ days of continuous flight. Still 81 hours is a long time. Looks like the limiting factor for low altitude systems has more to do with the weather than the available power. The high altitude systems can just fly over all the bad weather.

After looking at their website & YouTube I think they used a nose mounted direct drive motor.

Here's their paper - they used 5:1 gearbox to get better efficiency from the motor.

 

[quorneng]

Keaton
I am still not convinced about the Prantle wing concept. If it really produced a 20% inprovement over a more conventional layout in a practical airframe I do wonder why it has not been more widely adopted in situations where aerodynamic effciiency is a prime concern.

For my experiment being just battery powered i was more concerned about minimum weight than achieving absolute aerodynamic afficiency particularly at its relatively small size the effects of a low Reynolds number would make the aerodynamics less significant than the weight.
I have been building planes using thin sheet foam (Depron) for several years and discovered than it can be used to create a monocoque structure where it carries virtually all the loads with just limited internal support to hold the required profile.
The plane configuration is a pod and boom fuselage, a pylon mounted wing with a pusher prop. althoug probably not the most effcient layout it did suit the positioning of the electrical compnents, gave the natural stability of a high wing layout and kept the majority of the airframe clear of the high(er) speed prop wash. It oriignally had a V tail but control issues meant a return to a convention layout. The wing centre section is flat with diheral tips

The big battery is in the pod to counter balance the weight of the motor and tail. The wing pylon is high enough so the prop (8x6) clears the
light weight glass fibre fishing rod. The pylon also contians the speed controller and receiver placing all the components in a logical connection sequence. The motor is a 24g 1300kV Blue Wonder.
Ready to fly plane weighs 436g of which the 5Ah battery makes up eaxctly half.
This shows the construction of the relatively highly stressed wing centre section.

All 2mm Depron sheet with 1.5 mm hard balsa top and bottom spar flanges over a Depron shear web. The outer wing panels are similar but do not need the balsa spar flanges.
The tiny 3.7g aileron servos are buiried near the wing trailing edge such that the servo arms are glued directly on to the aileron

The servo itself thus acts as the inboard pivot. The outboard is a simple pin pivot.
The round nose aileron is very carefully faired so in the normal flying position there is the minimum of drag and air leakage bottom to top.

It creates a very 'clean' wing tip. A important consideration given its generous wing area and relatively small tail surfaces.

The elevator servo is buried in the fin. There is no tudder it is flown "bank and yank".
Although in theory it could fly for 5 hours all testing was done with a 1Ah battery with ballast up to the weight of the 5 Ah one. In this format repeated improvements and tests were made until it could fly for 1 hour meaning a 1A average consumption.
I never tried the full duration possible with the big battery but flying alone on a flat site I discovered that 2 hours was my limit rather than the planes. I had a stiff neck for a couple of days.
This was all done some 10 years years ago but with modern GPS gyro systems the need for human control input would be virtually eliminated.

Just an aside but the resulting airframe is strong enough for modest loop and roll aerobatics and it will not spin.

 

[Piotrsko]

Heed well the advice of @quorneng. He really knows his stuff. If youre making lift, you also make drag, proportionally.
Never had an issue with prop stall, going sonic somewhere near the tips is the usual issue even in helicopters with a 20:1 reduction on a meter diameter rotor. You'll hear sonic, one of the reasons drones are so annoying.

Can you thermal? How about slope? How are you flying this at night? I have a thermal source from the heat plant at the neighborhood school that is available oddly 24 hrs a day but you're looking at 3+ days and nights and either solar recharge or some massive radio battery.

 

[HVB79]

Here's their paper - they used 5:1 gearbox to get better efficiency from the motor.

From the paper:

The aircraft is driven by a RS-E Strecker 260.20 brushless DC motor with kV = 400RPM/V that works at up to 450W electrical input power. It drives an all-steel planetary gearbox with four pinion gears and a 5:1 reduction ratio and a foldable custom-built carbon-fiber propeller with diameter D = 0.66m and pitch H = 0.60m. The gearbox had to be upgraded from previous designs to increase propulsion system reliability during multi-day flights.

Why not use the same obscure motor and a planetary gearbox from the same company as a starting point?

RS-E Strecker 260.20:

Strecker-RS 260.20... 8/10 oder 14pol-Motoren - Welle 3,17 + 4mm

Reisenauer Micro Edition 3,5:1 planetary Gearbox:

Reisenauer Micro Edition 3,5:1Planetengetriebe

 

[Keaton Hare]

Keaton
I am still not convinced about the Prantle wing concept. If it really produced a 20% inprovement over a more conventional layout in a practical airframe I do wonder why it has not been more widely adopted in situations where aerodynamic effciiency is a prime concern.

For my experiment being just battery powered i was more concerned about minimum weight than achieving absolute aerodynamic afficiency particularly at its relatively small size the effects of a low Reynolds number would make the aerodynamics less significant than the weight.
I have been building planes using thin sheet foam (Depron) for several years and discovered than it can be used to create a monocoque structure where it carries virtually all the loads with just limited internal support to hold the required profile.
The plane configuration is a pod and boom fuselage, a pylon mounted wing with a pusher prop. althoug probably not the most effcient layout it did suit the positioning of the electrical compnents, gave the natural stability of a high wing layout and kept the majority of the airframe clear of the high(er) speed prop wash. It oriignally had a V tail but control issues meant a return to a convention layout. The wing centre section is flat with diheral tips
View attachment 240322
The big battery is in the pod to counter balance the weight of the motor and tail. The wing pylon is high enough so the prop (8x6) clears the
light weight glass fibre fishing rod. The pylon also contians the speed controller and receiver placing all the components in a logical connection sequence. The motor is a 24g 1300kV Blue Wonder.
Ready to fly plane weighs 436g of which the 5Ah battery makes up eaxctly half.
This shows the construction of the relatively highly stressed wing centre section.
View attachment 240323
All 2mm Depron sheet with 1.5 mm hard balsa top and bottom spar flanges over a Depron shear web. The outer wing panels are similar but do not need the balsa spar flanges.
The tiny 3.7g aileron servos are buiried near the wing trailing edge such that the servo arms are glued directly on to the aileron
View attachment 240324
The servo itself thus acts as the inboard pivot. The outboard is a simple pin pivot.
The round nose aileron is very carefully faired so in the normal flying position there is the minimum of drag and air leakage bottom to top.
View attachment 240326
It creates a very 'clean' wing tip. A important consideration given its generous wing area and relatively small tail surfaces.
View attachment 240327
The elevator servo is buried in the fin. There is no tudder it is flown "bank and yank".
Although in theory it could fly for 5 hours all testing was done with a 1Ah battery with ballast up to the weight of the 5 Ah one. In this format repeated improvements and tests were made until it could fly for 1 hour meaning a 1A average consumption.
I never tried the full duration possible with the big battery but flying alone on a flat site I discovered that 2 hours was my limit rather than the planes. I had a stiff neck for a couple of days.
This was all done some 10 years years ago but with modern GPS gyro systems the need for human control input would be virtually eliminated.

Just an aside but the resulting airframe is strong enough for modest loop and roll aerobatics and it will not spin.

That looks amazing! The servo as a pivot point is highly ingenious - I will have to look into if it will work on our model. A 5 hour endurance is really impressive for such a small plane! Regarding Prandtl - Albion has done tons of prototyping of prandtl gliders to prove efficiency - It is limited by the difficulty to design and lack of funding at the moment, as there are four people on earth who can design one properly. he and his team of interns built around 30 prototypes to study efficiency and handling (the biggest had a 27ft wingspan), and a prandtl derived prop was tested at a 20% increase in thrust, and a corresponding decrease in noise, versus a traditional prop on the same motor. There are actually prandtl props on all fans in the ISS to reduce noise, designed by Albion. He was chief scientist at NASA Armstrong for 20 years and this was his life's work essentially, and I have an excellent chance to prove its efficiency. At the least, however, it does provably create an inherently stable flying wing, which definitely provides a 10-20% decrease in drag purely from the removal of all surfaces besides the wing (plus killing the induced drag of the horizontal stab). We intend to use balsa ribs and a carbon spar for minimum weight - Ideally we will have an all-up weight of less than 10-15 lbs, 8 or so being battery mass (at the moment i plan to use 3c with Molicel p45b LION cells.) I would love to have you onboard if you'd be willing to advise, we are in the final stages of CAD and I definitely would appreciate some expertise in that department. Ideally we are going to attempt a HAPS (high-altitude psuedo satellite) with a 27ft wingspan and hopefully greater payload. I would love to talk further on zoom or something - you seem highly experienced and I would love your help. Shoot me an email at keatonhare@gmail.com if you have time!

 

[Keaton Hare]

From the paper:


Why not use the same obscure motor and a planetary gearbox from the same company as a starting point?

RS-E Strecker 260.20:

Strecker-RS 260.20... 8/10 oder 14pol-Motoren - Welle 3,17 + 4mm

Reisenauer Micro Edition 3,5:1 planetary Gearbox:

Reisenauer Micro Edition 3,5:1Planetengetriebe

I believe they custom made their gearbox but honestly that might be the best way to go - I hadn't noticed Reisenauer had a 5:1 that's super useful!

 

[Keaton Hare]

Heed well the advice of @quorneng. He really knows his stuff. If youre making lift, you also make drag, proportionally.
Never had an issue with prop stall, going sonic somewhere near the tips is the usual issue even in helicopters with a 20:1 reduction on a meter diameter rotor. You'll hear sonic, one of the reasons drones are so annoying.

Can you thermal? How about slope? How are you flying this at night? I have a thermal source from the heat plant at the neighborhood school that is available oddly 24 hrs a day but you're looking at 3+ days and nights and either solar recharge or some massive radio battery.

I'm a glider pilot so that definitely influenced how I approached this - We will be using ardupilot which can soar autonomously, but for our purposes we need to stay aloft in still air or even sink through 24 hours. We will be using 180000ish Mah worth of Molicel P45b for the battery and sunpower E60 across the wing for solar recharging. We don't know the final powertrain so it is hard to estimate endurance, but minimum 8-10hrs of battery only flight, and maybe more if we can get it. In the final high-altitude version we will be 20k-60k up so thermals won't be super useful, but I might build in some wave pathfinding stuff to take advantage of mountain wave if we can get it.

 

[Piotrsko]

Last I heard you need lots of altitude for flying mountain wave, and it's unpredictable. Then there's the stall speed vs max airspeed issue which is the same for your possible altitude choice. Flutter or stall is a bad choice to have. Col Payne had, IIRC, 2 knots slop between stall and DNE when he did his fast long record breaking flight ten years back.

 

[L Edge]

Is this a college or Darpa project demo?

 

[Keaton Hare]

Is this a college or Darpa project demo?

Neither, although most of my team are undergrads. I'm doing it as a passion project and hopefully to help prove Prandtl wing's efficiency.

 

[Keaton Hare]

Last I heard you need lots of altitude for flying mountain wave, and it's unpredictable. Then there's the stall speed vs max airspeed issue which is the same for your possible altitude choice. Flutter or stall is a bad choice to have. Col Payne had, IIRC, 2 knots slop between stall and DNE when he did his fast long record breaking flight ten years back.

Wave is not terribly hard to use in a glider. However, with a UAV it is slightly harder, but still not impossible. Wave is typically between 10,000 and 30,000 feet - our second prototype will ideally have a wingspan of about 27 feet and operate between 20,000 feet and 60,000 feet. Wave is not impossible to predict as it happens when wind at a speed of 15-ish knots on a normal day, and even less on a very smooth day, intersects a mountain range at about 90°. With such a light aircraft, I do expect to run into some flutter issues, but I think it can be effectively resolved with the right air frame construction.

 

[Keaton Hare]

Wave is not terribly hard to use in a glider. However, with a UAV it is slightly harder, but still not impossible. Wave is typically between 10,000 and 30,000 feet - our second prototype will ideally have a wingspan of about 27 feet and operate between 20,000 feet and 60,000 feet. Wave is not impossible to predict as it happens when wind at a speed of 15-ish knots on a normal day, and even less on a very smooth day, intersects a mountain range at about 90°. With such a light aircraft, I do expect to run into some flutter issues, but I think it can be effectively resolved with the right air frame construction.

Take the above explanation with a grain of salt as even though I know how to use it, I am definitely not qualified to explain it.

 

[Piotrsko]

Psel with a glider endorsement.. No need to explain wave, how to predict / find / use, just know it's harder than the books indicate from experience. Or I am a lousy pilot. Or both.

Otoh, once every so often, we get someone like you who wants to do a duration or altitude remote aircraft. Never get to see the result of their efforts come to fruition.. at least you don't need a pressure suit.

 

[quorneng]

Autonomous flying at 30.000+ft.
Surely you will need the authority from the FAA and have to meet stringent collision avoidance criteria for each and every flight if flown in the USA.