Airfoil selection for perpetual 5m solar eGlider

[WulfKeeper]

On the topic of flight profile. This is a grab from the original Solar Impulse paper, "“Fly around the World with a Solar Powered Airplane" by Hannes Ross:

It's a very interesting read!

 

[Piotrsko]

Chip in's most welcomed!

Did you spot weld the cells? I'm a bit reluctant to go down that path... I *think* I can do it without killing myself

There are diy data out there to print battery holders and there are ready to assemble packs for all the variants. Spot welding is merely annoying repetition once you work out your safety backup parameters. Might want to get batteries sorted first since pouch batteries don't need spotwelding ad the differences affect most of the other planform variables

 

[Jackson T]

Very interesting! I was a bit confused for a minute that they were considering flying AtlantikSolar at such high altitudes, then I went to the top of the post and realised that it was the Solar Impulse .

 

[WulfKeeper]

Very interesting! I was a bit confused for a minute that they were considering flying AtlantikSolar at such high altitudes, then I went to the top of the post and realised that it was the Solar Impulse .

Yeah, and down here in Oz we're offically limited to (IIRC) something like 120m? Telemetry feed range should at least be easier. LoRa might be fun..

I wonder how hard it is to get permission to go above 120m if I'm prepared to drive out into the middle of nowhere... Another 'Hmmm'.

 

[WulfKeeper]

There are diy data out there to print battery holders and there are ready to assemble packs for all the variants. Spot welding is merely annoying repetition once you work out your safety backup parameters. Might want to get batteries sorted first since pouch batteries don't need spotwelding ad the differences affect most of the other planform variables

Yep. Inside a (conductive) CF spar only adds to the 'fun' too! The 3D printed holders offer significant appeal for this reason too (insulation).

 

[Piotrsko]

Yep. Inside a (conductive) CF spar only adds to the 'fun' too! The 3D printed holders offer significant appeal for this reason too (insulation).

Isn't conductive under all regimens. They do put batteries in the metal frame tubes of bicycles which only occasionally catch fire (but not by shorting out).

 

[quorneng]

WulfKeeper
I actually used ordinary cells and used the box spar as a battery holder with a centre spring loaded contact.

The plane had a 40" span and was light so it only pulled 5A at full power and cruised on just 2A (12W).
Although there was some concern at the proposal, it was pointed out that some clubs will not allow spring battery contacts on anything, it actually worked very well and was completely reliable.

 

[WulfKeeper]

WulfKeeper
I actually used ordinary cells and used the box spar as a battery holder with a centre spring loaded contact.
View attachment 212878
The plane had a 40" span and was light so it only pulled 5A at full power and cruised on just 2A (12W).
Although there was some concern at the proposal, it was pointed out that some clubs will not allow spring battery contacts on anything, it actually worked very well and was completely reliable.

Thanks for sharing quorneng. Nice approach! Simple and obviously effective. I'm toying with something similar myself. 'Button' 18650 cells, single end-to-end, under some type of compression/hold in the wing spar. Arranging the 6s3p (and a centre 8s3p) configuration/wiring is also going to be 'interesting'!

One of the things I'm really enjoying about this project is the number of different areas of 'design challenge'! Aero theory, design, construction techniques, electrics, electronics etc etc. Keen or stupid? I dunno

 

[L Edge]

Thanks for sharing quorneng. Nice approach! Simple and obviously effective. I'm toying with something similar myself. 'Button' 18650 cells, single end-to-end, under some type of compression/hold in the wing spar. Arranging the 6s3p (and a centre 8s3p) configuration/wiring is also going to be 'interesting'!

One of the things I'm really enjoying about this project is the number of different areas of 'design challenge'! Aero theory, design, construction techniques, electrics, electronics etc etc. Keen or stupid? I dunno

Interested in info on how you are going to figure out your system wattage output dealing with latitude, time of year, length of day, angle of inclination of rays on your airfoil(many others factors). What control zone for flying are you in(FAA or other)?

 

[WulfKeeper]

Interested in info on how you are going to figure out your system wattage output dealing with latitude, time of year, length of day, angle of inclination of rays on your airfoil(many others factors). What control zone for flying are you in(FAA or other)?

G'day L Edge

Short version: like all other RD&D projects! Literature review (e.g. including the excellent solar impulse and Atlantik Solar papers), grab some theory, run some sims / numbers. Get a vague idea about how it looks and might perform. Then incrementally Real World test where I can (will rig up 'sims' on the ground, PV arrays, motor, MPPT etc etc ), do lots of flying and run some V&V.

With the very clear understanding that observations trump theory by a *lot*! Enormous parameter space and it'll swing wildly

 

[quorneng]

WulfKeeper
I think you may find that high altitude (40,000+ ft) is chosen more for smooth air conditions above the 'weather' than solar panel efficiency. The extra weight to provide strength for handle atmospheric turbulence reliably is significant the bogger the plane becomes.

 

[Jackson T]

Yeah, and down here in Oz we're offically limited to (IIRC) something like 120m? Telemetry feed range should at least be easier. LoRa might be fun..

I wonder how hard it is to get permission to go above 120m if I'm prepared to drive out into the middle of nowhere... Another 'Hmmm'.

I thermalled my 4m balsa glider to 300-400m altitude a little while ago, but I'm way out in country NSW. Where abouts are you based?

 

[Jackson T]

WulfKeeper
I actually used ordinary cells and used the box spar as a battery holder with a centre spring loaded contact.
View attachment 212878
The plane had a 40" span and was light so it only pulled 5A at full power and cruised on just 2A (12W).
Although there was some concern at the proposal, it was pointed out that some clubs will not allow spring battery contacts on anything, it actually worked very well and was completely reliable.

That is VERY low power, nice work! How heavy is it? Also, why are spring contacts used instead of conventional contacts?

 

[quorneng]

That is VERY low power, nice work! How heavy is it? Also, why are spring contacts used instead of conventional contacts?

The Battery Wing weighs 10 oz (284 g) of which the four 18650 cells account for 4.8 (136 g)!
The cells had no tags so a spring contact was the only option.
This is its final MkV form.

Although at the time I was concerned at the effect of roll inertia from the heavy cells spread out along the wing it was actually the yaw inertia that proved to be the problem eventually requiring a much bigger fin and some fuselage side area to achieve an adequate lateral stability for comfortable 'endurance' flying.
The very poor quality of the 18650 cells I used (I later found out they were rebranded rejects!) actually limited its endurance to about 40 minutes. With proper 2000 mAh cells it could have managed 2 hours but they would have cost quite a bit more and be significantly heavier than a similar capacity LiPo.
Now that 18650 cells have improved capacity and LiPo prices are rising it might be worth revisiting the 'battery wing' concept.

 

[WulfKeeper]

The Battery Wing weighs 10 oz (284 g) of which the four 18650 cells account for 4.8 (136 g)!
The cells had no tags so a spring contact was the only option.
This is its final MkV form.
View attachment 213418
Although at the time I was concerned at the effect of roll inertia from the heavy cells spread out along the wing it was actually the yaw inertia that proved to be the problem eventually requiring a much bigger fin and some fuselage side area to achieve an adequate lateral stability for comfortable 'endurance' flying.
The very poor quality of the 18650 cells I used (I later found out they were rebranded rejects!) actually limited its endurance to about 40 minutes. With proper 2000 mAh cells it could have managed 2 hours but they would have cost quite a bit more and be significantly heavier than a similar capacity LiPo.
Now that 18650 cells have improved capacity and LiPo prices are rising it might be worth revisiting the 'battery wing' concept.

Yeah, there's a *huge* variation in the quality of the cells. Even the different revisions (latest tech) across a respected brand can see significant improvements in cell performance.

I'll definitely be spending some time picking the best flavour I can afford before forking out the readies!

I thermalled my 4m balsa glider to 300-400m altitude a little while ago, but I'm way out in country NSW. Where abouts are you based?

G'day Jackson

I'm on the NSW Central Coast. Locally, I'm pretty much stuffed. I've got a hospital helipad up the road and I'm within 15km of the local aerodrome. I can do some basic 'will it fly' tests but nothing too high. Coastal soaring isn't really applicable to this gig either

If I head north I run into RAAF restricted airspace with Williamtown. South puts me closer to YSSY... So 'go west'!

I reckon the other side of the Great Divide would be the best idea. Maybe Bathurst, Orange, Mudgee? I've got some friends of friends I might be able to impose upon but they're not ag-type folk. I've got some mates out near Narrabri that might be worth hitting up. I don't mind the drive and it's lovely out there (depending on the time of year).

Pity I don't know anyone at, say, Breeza. Vast tacts of huge paddocks of various crops (mainly wheat). Poor b@stards got flooding at just the wrong time this year.

Anyway, somewhere like that'd be near on perfect for this type of flying Open to suggestions (as always)

 

[Jackson T]

The Battery Wing weighs 10 oz (284 g) of which the four 18650 cells account for 4.8 (136 g)!
The cells had no tags so a spring contact was the only option.
This is its final MkV form.
View attachment 213418
Although at the time I was concerned at the effect of roll inertia from the heavy cells spread out along the wing it was actually the yaw inertia that proved to be the problem eventually requiring a much bigger fin and some fuselage side area to achieve an adequate lateral stability for comfortable 'endurance' flying.
The very poor quality of the 18650 cells I used (I later found out they were rebranded rejects!) actually limited its endurance to about 40 minutes. With proper 2000 mAh cells it could have managed 2 hours but they would have cost quite a bit more and be significantly heavier than a similar capacity LiPo.
Now that 18650 cells have improved capacity and LiPo prices are rising it might be worth revisiting the 'battery wing' concept.

So is it no good to solder a lead on to the ends of the cells?

 

[quorneng]

A 18650 cell case is made of stainless steel. Normal solder will not touch it.
There are special solders and flux that can solder SS but on a cell you would have to apply so much heat to get the required temperature it would damage the cell. The tags, which are pre tinned, are spot welded on and are thin enough to allow them to heat up for soldering without transmitting too much heat to the cell.
If it needs to be soldered connections it has to be tagged cells.

 

[WulfKeeper]

Ok back to a bit of design theory... ailerons

Length
So theory says ~0.3 of wing length from fuse. Roughly 5m / 2 = 2.5 m
2.5 * 0.3 = 0.75 m
750mm / 125mm (per cell) = 6 cells works out about right. Bewdy.

Depth
Theory says 0.25 chord
Chord is roughly 300mm so call it 75mm.

Now that's neatly half a PV cell. Splitting them can be done (see the French solar e-glider's fruit ninja special on YT )
Or...
Go the SolarAtlantik route and use a full cell (125mm) but severely curtail the throw.

Current thought is to go with the full cells. Maybe having all that control surface influence on call might come in handy for inducing some roll in this beast?! (Refer earlier in the thread to the 'just let the FC software figure it all out' philosophy )

Some inbuilt dihedral (7deg) is back in the design. Not convinced I'm going to get the friendly flex I desire out of the wings.

Test flights, without the true ($ exe) PVs, remain a solid part of the plan for this sort of reason - get the airframe sorted first.

Yes, I fully expect to get this wrong... just saying... again. 'Iterations' is my word dejour!

Thoughts / comments welcomed, as always.

W

 

[quorneng]

Solar cells on the ailerons.
You need to be careful about the weight added to a moveable surface or you could induce flutter particularly on a light and thus a relatively flexible wing. Any 'give' in the linkage or even flexibility of the aileron itself can induce a sympathetic oscillation in the wing which can 'auto develop' to wing failure.
Don't worry about the 'theory' as that refers to an 'average' plane which clearly yours will not be. Once you move out of the need to perform any sort of aerobatics the control surfaces areas can be reduced, particularly for the ailerons.
The roll effectiveness of an aileron is greatest at the wing tip so a small broader area close to the wing tip will be just as effective as a longer thinner one. It is also likely to be more rigid for the same weight.

As an example just look at the size/area of the aileron on a Bombardier Q400 (marked in red).

I built an scale RC one and yes its roll rate is modest and it would likely be a handful in very turbulent conditions but I don't fly it in those sorts of conditions and I don't expect you would either.

My own suggestion would be to reduce the area of the ailerons as much as possible (by testing) and certainly make them a suitable size for any solar cells. Its not as if you will be using the ailerons much at all. My endurance flying consists of just 'nudging' the plane's direction to keep it within a suitable visual distance.

 

[WulfKeeper]

Been running some more design numbers. Comparing to other airframes. One thing that really stands out is the disproportionately large tail fin / vertical stabiliser on the Solar Impulse versions. Roughly 1.5x taller than theory suggests they need to be (in a nutshell if you cut them off where they start to extend below the fuse then it's about right... and they extend quite a way below...).

No clear explanation of the design decision I can readily find...

I'm thinking supreme levels of rudder command, for flat turns to help keep the panels pointing up?!