Wish I had a large surface for you.
Sorry, I forgot to get back to you on that point.
I am familiar with the uBLOX modules. I used these instructions to put it into flight mode, and it appears to behave as expected (but we haven’t done another balloon test to verify). They supposedly work up to 164k feet in that mode.
I’m not sure the same can be done with any integrated GPS device–it typically requires somewhat low-level access. As you can see from the link, in the uBLOX case it requires sending some very specific strings, which likely can’t be sent with any unit that’s built in. So I suspect you’ll have to find a way to integrate one of these units to your setup (they are at least cheap: like $10 on eBay).
I think it would be fun to build an autonomous glider that is hauled aloft to the edge of space by a balloon. The glider would have a whole suite of sensors, including cameras. The balloon would attach to the tail of the glider. As the balloon ascended, the glider would monitor it’s position, altitude, rate of climb and time. If the glider sensed that it was at a predetermined distance downrange, above a predetermined altitude, that the balloon had burst, or that it was simply time, it would detach from the balloon and begin to glide back to a predetermined landing position and land itself.
680 kg, according to other pages. But that thing is monstrous.
Consider just the size needed for a balloon to lift its own mass, remembering that at 100k feet, the pressure is about 1% of sea level:
(4/3pir^3) / (2.24 m^3/mol @ 0.01 atm) * ((29-2) g/mol) = (4pir^2) * (0.0000381 m thickness) * (926000 g/m^3)
I used the 1.5 mil thickness from the NASA link and a density of 926 g/cm^3. Solve for r and you get 8.8 meters. That’s huge! And just for zero payload. The balloon weighs about 35 kg.
Still, you can do better with thinner plastic. It’s been done–the people I worked with have done them before, with tiny payloads (<20 grams, IIRC). But you aren’t putting a GoPro on one unless you have pretty serious resources.
If you use one of those, you could have an indifferent flight!
I’m pretty sure that they meant “indefinitely”. 
I’m not sure, but the FAA might have something to say about that.
The smaller radius means I could use thinner (less strong) LLDPE, but I don’t know how readily available that is so at best I might have to use 1.5 mil LDPE, IIRC that’s only a little lighter. So it would get big, but either way I probably can’t get large enough sheets which means seams which means too much build time. Except it’s so tempting. Heat sealing seems impractical structurally, maybe tape reinforcement and cheap heat sealing to prevent leakage. I’ll never get that done alone.
I took a glance at the Garmin Virb and that looks great if I can find a way to transmit the data back. What did you find out about using GPS above 60K?
Ok, so the equilibrium point comes from the total weight and maximum volume of gas at some altitude where lift drops to zero. The curvature of the earth can be seen at 33,000 feet, but just barely and probably blocked by clouds. But maybe I could try for 60,000 feet where GPS still works. In theory it could release gas if it exceeds some altitude, like 60,000 feet if I use the GPS instead of air pressure to determine altitude.
BTW, here’s a map of data points from our launch. Altitudes are in meters.
You can see that the data looks pretty good at first. There are a couple of cutouts, but if you follow the path you can see that we get good data up until 12,000 meters. Over that point, the data goes completely crazy, although oddly it still stays within a few hundred miles. It seems like they’re adding an intentional jitter above the limit.
So I misremembered acutally; our unit cut out above 12k m (40k ft). Regardless, the unit would have been fine with the hack.
We had a barometer up there as well. It went totally crazy when the pressure got too low. We should have tested it in a vacuum chamber first. Not that it mattered much since it wasn’t designed to be accurate at those low pressures.
Thank you, there is a larger percentile differential in lift that benefits Hydrogen over Helium that I remembered.
Cost was not a major consideration per the OP, and flammability was. I was unaware that these parameters had changed.
This notion is debatable…
Here’s my suggestion for your problem. It requires a bit of electronics and programming:
- Get a cheap uBLOX-6M module from eBay
- Get a cheap bluetooth<->serial adapter from eBay
- Get them hooked up and pair with phone
- Instead of using the phone GPS positioning, connect to your bluetooth adapter and read the serial data directly (NMEA format). You can use the sample code from here where I used the same bluetooth adapter.
- Also, use the instructions I linked to above to put it in flight mode (i.e., supports >60k ft)
It’s a bit of extra effort but basically guaranteed to work. As best I can tell, the built-in GPS units on phones are basically guaranteed not to work.
End of the week wrap-up.
Thanks everyone for providing great input! Shout outs to scr4, Dr. Strangelove, and Spiderman, and appreciating all other responses as well.
Communications:
Cell phone out, satellite phone in. Yeah, they’re more expensive, but the prices have come down, and maybe I can find something that’s just a data communication device. I don’t know for sure they work at high altitude, don’t see why they wouldn’t.
Couple of options:
Irridium: Global coverage, more expensive.
Global Star: Weak coverage in the arctic and mid-ocean.
GPS:
They are limited to altitude < 18,000 m or velocity < 515 m/s. Dr. Strangelove provided a way to hack that. It would violate import/export laws to have a hacked unit cross an international border. I’m not worried because: A) I’ve worked as an international technology smuggler before B) It may be both altitude AND velocity restricted, so maybe the hack can be limited to just altitude.
Batteries:
It’s gonna get cold up there. The electronics really need to power down between readings and stay operable in the extreme cold.
Balloon:
Several problems.
Zero pressure requires carrying ballast and/or extra gas. A single release of gas or ballast to isn’t that big a deal, but a complicated active system seems difficult. Not ruling it out yet.
Superpressure doesn’t seem that bad. Lot’s of technical info online. The big problem is the large size of strong material. That makes it heavy, and it will need seams that don’t leak. It doesn’t look as bad as I thought, but it will be very large because of the heavy material and limitations on the ability to make the seams. I have an idea that will bring this close to practicality, almost close enough to do. I’ll detail later
Complexity:
This is always the bugaboo. Electronics are building up, satellite phone, GPS device, camera, batteries, extra camera, release device, and maybe a cell phone still needed for interfacing, and an external power controller. All this adds weight, feasible balloons are heavy too. This has to be done with hydrogen, the cost of helium for something this heavy would be too high. It will also require intermediate test stages before I spend this much all the way through. I think my balloon idea makes this more realistic by keeping the cost down.
Iridium is fun; I have some experience with that, too.
They will work at high altitude. They will even work in orbit! We’d actually intended to use an Iririum unit for one of our cubesats, though the idea fell through. It’s been done before, and problems are more regulatory than technical. For a balloon, I don’t see there being a problem.
My experience is with the SBD (short burst data) modems. They’re lightweight and easy to interface with. I can get you some code and other information if you’re seriously considering this. Comms are bidirectional, too, so you can query the balloon and change its behavior.
Obviously they’re quite expensive per-byte–something on the order of a dollar per kilobyte. For position tracking and such, that’s not so bad. A bit high for imagery, though.
As for batteries, we used non-rechargeable lithiums (Energizer brand). They worked fine down to -40 C or so. The (experienced) guy we worked with said they were the right thing to use. I dunno what you’d use if you want rechargeable, though.
I wasn’t expecting to charge the batteries, though it might be interesting to carry some lightweight solar cells up just to see how much they can produce in the thin clear air. The little tracker devices are cute, maybe what would run most of the time, with pictures taken sparingly and sent by sat phone, or held on a cell phone on the chance it can send them during a final decent or on landing. Don’t need a lot of pictures, but after putting a lot of work into something I probably will never see again I’d like to get something back more substantial than just coordinates. If I get one picture that shows the curvature of the earth clearly, and something straight down that isn’t just clouds I’ll be pretty happy.
Before long I’ll reach out to the HAM community and see if there’s an alternative to SAT phones.
One thing to look into is the IR absorption by hydrogen, helium is apparently invisible and only the surface area of a balloon contributes to heating. If the volume of gas is soaking up IR then I’ll need a stronger balloon.
A 10 KB 320x240 image doesn’t look too bad, and only $10 to download over Iridium. 1600x1200 @ 100 KB looks pretty reasonable as well, and $100 might be worth it for a particular “money shot”. So you have some options here.
As far as the Ham stuff goes, there are certainly big networks for position reporting, and you can accomplish this with very low-powered transmitters. I doubt they can handle any reasonable amount of data, though. I’m a bit weak on this end of things, though, so it’s worth diving into more.
I just realized that the group we worked with has done a bunch of superpressure balloon flights, including some record-settings ones. Here’s their website. They have a little news item on our flight as well, “CNSP and Southern Stars”. I’m the guy in the picture with the Yagi antenna and laptop :).
Another variation of your effort.
Stuffed dog and a balloon.
How about a DeLorme InReach? Built-in GPS, tracking function, and can be used to send text messages - all via Iridium satellite system.
I’d strongly suggest that you aim to track the balloon and recover the payload. Transmitting all collected data during the flight will be difficult.
You’ll thus need a scheme for ending the flight and “de-orbiting” your payload. This could probably be done by causing the balloon to deflate and become a “streamer” that yields a fairly rapid rate of descent but still slow enough that your small payload can survive. Might want to figure out how to jettison the batteries (perhaps one-by-one, as they expire? - shedding weight could prolong the flight).
Onboard smarts could be via an Android smartphone, or something like a Rasberry Pi / Beaglebone / Arduino.
Building a Superpressure Balloon
I was getting a little discouraged about this subject. **scr4 and Dr. Strangelove **mentioned using thinner plastics, but cost, availability, and workability are big factors and I had my concerns. After some more research things don’t look so bad. Just a little more complicated than I would prefer.
The main issue is tensile strength of the material. Superpressure means constant volume so as the gas attempts to expand as the differential between internal and external pressure increases the material can’t stretch much, and can break under stress. LLDPE and Mylar seem to be good alternatives, but I haven’t seen LLDPE in anything greater than 48" widths, and only a little wider for most Mylar suppliers, but I recall Mylar is made up to at least 96" width, though maybe not the thin material I’d need. Mylar has the additional problem of sealing, it sticks better than xxPE to adhesives, but it can’t be heat sealed (melting temperature changes the molecular structure of the plastic making it brittle).
Then assuming I can find the right material this is too big for any single sheet of anything, even 40 foot wide LDPE, so it will need seams. Balloon seams are a complex subject. The seams have to be structurally sound and seal in the gas. Sealing is tough for any method, hydrogen and helium both leak through the tiniest of pores because they are such puny little atoms.
Heat sealing makes the best seal. Imperfect for any single continuous seam, but folding the material and increasing the width of the sealing area will reduce the chances of a leak, and then potentially combining that with silicone adhesive in the joint or silicone tape over the joint will have to be good enough. Heat sealers are generally constant heat or impulse strip sealers, or constant heat irons, and constant heat rollers. Used strip sealers can be found for reasonable prices or built, as can rollers. Small irons are cheap enough to buy, and in theory a clothes iron will do with a variable power source (damn! I wish I never sold that big ass Variac I had).
Seams also have a few different configurations. The best is the overlap, a couple of inches of material in each piece flat on top of each other. The other main variety of seam is the standing seam where as typically used in party balloons and where the joined material sticks out perpendicular to the material on either side, like the post of T.
Standing seams are much easier for various reasons, especially closing seams such as joining up several pieces to make a tube. Pressure has to be applied from both sides so when you close things up with overlapping seams you run into topological problems trying to do that with huge pieces of material. On top of that you need to lay the material side by side, the logistics are daunting for big pieces of thin material. There’s a big disadvantage to standing seams though, they are weak. The tension on the seam tries to peel it apart, in a cross section the forces reduce to a point. An overlapping seam is pulled in the plane of the joined material, which would be inches wide. It’s the difference between peeling off one of those annoying stickers that come on everything you buy, or sliding the sticker off. Both types of seams can be further refined by folding.
For practical purposes standing seams will be needed. A wide seam would be heat sealed or joined with silicone adhesive strips (they come on a roll with a backing strip that is peeled off leaving just the silicone adhesive, and in both room temperature and heat setting forms). The joined material can be folded and the folds sealed to strengthen the joint and provide better stress relief, and potentially stitched past the initial seal for fail safe protection and to prevent a small seperation from expanding. In theory the material can be passed through a guide which folds the material together for a single pass to include the folds. I doubt that will be practical, handling thin plastic is already enough of a challenge.
Materials are going to have to be something available in widths of at least 10 feet or there will be just too many seams to deal with. What’s available and makes sense are LDPE in various forms and thicknesses, HDPE. If I can find wider Mylar I’ll look closely at that, it’s a strong and has minimal porosity even in very thin material. I have to look into HDPE physical characteristics. It is available in thicknesses like .3 mil. It’s considered stronger than LDPE, but that’s because it stretches much further before breaking, doesn’t sound good for a constant volume balloon. It’s also opaque white, somewhat reflective, but it’s going to get a lot hotter than clear material. Heavier materials weigh more, which makes the balloon larger, but thicker and heavier materials are easier to handle, and cost is neglible for LDPE. I’ll consider using something like 3 or 4 mil LDPE because it’s available in larger sizes and should have plenty of strength.
I was still worried about the shape of the balloon, spheres are out of the question, cutting gores in huge pieces of plastic is difficult and non-straight seams are a problem too. I thought of a couple of a few alternatives, and was thinking they wouldn’t work, but then I found this .pdf with lots of great top secret state of the art information when I was in the 1st grade. They performed tests using alternative shapes that I considered, the simple cylinder with gathered ends, sort of like a big joint on end, and the tetrahedron. They also ruled out short, wide onion shaped cylinders. They didn’t try something similar to the tetrahedron that I’ll just call the pillow case, one big wide sheet of plastic folded over with the sides seamed and the open end gathered. The corners would have to be reinforced in gathers, or additional seams use to curve the shape at the top.
Even forming gathers isn’t easy, the material has to be carefully pleated, folded over, the twisted and wrapped tightly. And they add weight in material that doesn’t contribute to the total volume, but they do help distribute the weight of the payload across all the material.
Their studies showed that the sphere was the most efficient shape, but tetrahedron and cylinders still worked, they’re just heavier and will result in bigger balloons made from heavier material. I’ll be testing the simple cylinder and pillow case, maybe the tetrahedron but it needs more seams, and as they note it there will be asymetric varying stresses on the material. The pillow case doesn’t distribute the load well but it’s got symetry to it and a minimal number of seams. They used Mylar for their tests. But that was in the early 60s. Every form of plastic has undergone considerable refinement in formulation and production since then, the physical properties are no doubt different, but hard to see where they wouldn’t be better.
So, I will move forward with testing. I can build small balloons in different materials and pressure test them to the point of destruction with air. That’s a small enough effort, but I can’t test them at low temperatures until next winter, which I won’t have the patience for. Maybe I can send small test bags to someone in Canada to try out, I understand it doesn’t go above 0C there until August (the C means Canadian I think, that’s 32 degrees American).
Small balloons won’t involve much build time. I can also make a full size balloon for pressure testing at trivial cost, but that will require more resources. It’s in the picture though, and before I’d spend big on the whole deal I’d have to go through these steps to know that I won’t just send something that will pop as soon as it gets to altitude. If I can feel secure that I have strong enough material and joints I’d make another one for a short flight test, designed to release the payload on a timer, which will parachute down, and carrying a cell phone so I can locate it when it lands. If all that succeeds I’d do the final build and send aloft all the equipment for the big journey.
Looks like a cool little device. Transmitting images in text packets is going to take a lot of packets though. I’d love to recover it, but I live on an island (sort of) and this thing could be out over ocean in no time, or worse yet,`the godforsaken interior of Connecticut. But yeah, I’d like a soft landing.
Dropping batteries as ballast is a brilliant idea! I just don’t know a simple way to do it, and each one would need a little parachute so I’m not endangering the public.But that’s part of the fun here, figuring out a way to do something like that. I love to add bells and whistles to things I build, my potato gun has a laser site and range finder!
Yes - that won’t work with the InReach. I expect that sending decent-resolution pictures via satellite is beyond what’s practical for this project.
You’d need to launch only when the wind is favorable.
Tracking info would tell you where the payload has landed.