All of the real nitty-gritty details are documented on our wiki which contains the real documentation and has much more detail. However, you might find it more difficult to navigate, sort of like reading Wikipedia from start to finish.

Go to the White Star Balloon Wiki Now

Here you can get a story-like description of which challenges we faced and the tech we used to meet them. Soon we’ll have a more systematic approach for those who already know a thing or two about ballooning.

the challenges

How do you…

1. make a balloon go where you want it to go?
2. keep a balloon at a desired altitude?
3. know where it is and what it’s doing?
4. ensure it will survive all the conditions it will encounter?
5. do this safely and legally?
6. solve all the problems above without adding too much weight?

1. getting the balloon to go where we want.

The balloon has no type of propulsion system, so what we have to do is get it into the jet stream and let it carry our inflatable, electronic friend across the Atlantic. The problem is the jet stream doesn’t always pass over Kentucky and make a straight line toward
Europe or Africa. So four times each day we download the Global Forecast System data from NOAA which is about 2GB. This, in itself, presents a challenge as bandwidth, storage, and processing power required adds up quickly.

This GFS data is processed by atmospheric dispersion modeling software typically used to track the path of air pollutant particles that result from traffic, industrial activity, or accidental releases. Our balloon, being adrift in the atmosphere, behaves more or less like a giant particle. The software spits out a mathematical model of the predicted path of the jet-stream layers.

The colored lines represent the different parts of the jet-stream. Each tick mark indicates a 6 hour period. Our balloon can maintain an altitude within the ~2 mile thick current of air for only about 36 hours so we need everything to line up and take a path that carries the balloon across the Atlantic within that time span. We only may have only a day or two notice on when to launch so we must be ready to break into action once a good pattern is predicted within a certain degree of certainty.

Tyler, a grad student at University of Louisville’s Speed School of Engineering, is our resident weather expert.

2. maintaining a desired altitude.

This is the story of how the balloon maintains a stable altitude and, for many, the most interesting or “OH, I get it” part of how the balloon works and why it’s not just a regular old weather balloon. As mentioned, the jet stream is a current of air only about 2 miles thick. We’re rising up to 35,000 feet, so this is a relatively narrow target. How do we keep the balloon from rising out of it and, later, sinking below it?

Upon launch, the balloon will contain more helium than it needs to stay aloft in order to rise quickly to about 35,000 ft. To stop itself, it must dump the excess helium.

So how does the balloon know how much helium to get rid of? Well, it doesn’t, but the valve sort of does. This balloon is a “Zero-Pressure” design. As altitude increases, air pressure decreases. This is what causes regular balloons to pop, the pressure inside the balloon envelope becomes higher than outside air pressure and it begins to expand until, eventually, the envelope breaks. The valve is designed to start releasing the gas inside before pressure causes the envelope to be stressed, thus achieving zero pressure. It does this with a simple spring. If pressure inside the balloon is greater than outside, the spring will compress and release some gas until the pressure is about equal. We’ll launch at night, after it first stabilizes it’s altitude, it should stay there… until the sun rises.

When the sun comes up, the gas inside the balloon will begin to heat up and rise again. For this reason, it is important that the valve allow helium to escape, but not allow outside air in because this extra volume will also expand. At night, the ballon should maintain a partial vacuum. It’ll simply release helium as before to avoid rising out of the jet stream. When night comes again, however, that same amount of gas will cool and it will begin to sink. We can’t add more helium, so we need to lose some weight.

The balloon carries three liters of liquid which it uses as ballast, if it needs to drop weight to rise, it drips out the liquid. The valve that releases ballast doesn’t work automatically like the air valve, it needs be able to make decisions… it needs a brain. Of course, this is an electronic brain and to make its decision it needs to know it’s losing altitude and carefully control the release of ballast and know when the altitude is corrected in order to close the valve. As you can imagine, this requires some programming. Check out the algorithm here.

We actually have another problem here. We used liquid because it flows out easily. There is no potential of jamming like you might have with a solid. However, when you’re traveling along at 35,000 ft, you’ll find that it happens to be around -40° and any liquids tend to, you know, become solid. Easy fix: we use ethanol which has a freezing point of -173°F.

As it travels through multiple sunsets and sunrises, it continues this process of dumping helium or ballast to sustain an altitude within the jet stream. It can do this for around 36 hours before we run out of helium and it crashes. This is why we need a jet stream that will complete the journey in less than that time.

3. communicating with the balloon.

This is definitely the part of our project that is the most difficult to explain since much of it involves lots of electronics and programming. Those well schooled on these subjects should consult the wiki for more detail, but here’s an overview.

It’s one thing to launch a balloon that can do all these things, but consider that you launch it and… well, you’ve launched it. Bye bye, what now? No, we need this thing to give us information and it needs us to do the same. The balloon needs to receive information from us to help it make decisions regarding it’s sinking/rising problems. We need to receive the data it is collecting, it’s location, altitude, and how it’s doing. Otherwise, this experiment is pretty useless.

When a two-way satellite text messenger for hikers hit the market at cheap price, we finally saw a way to hear from, and send instructions to, the balloon. We pulled the satellite modem out of the text messenger, and built our own circuit to control it. We can send it what it needs to make its decisions and it can tell us how it’s doing even when it is traveling over the open ocean away from cell phone towers and ham radio operators. We’ve gone even further and designed a shield for the Arduino microcontroller platform so others can more easily use this technology.

The end result is that we get a nearly real time report on altitude, global position, speed, temperature, pressure, and humidity (clouds). The fun part is that you do too. We’ve created an interface so anyone can track/receive the data just as we do. Be sure you’re signed up for launch notification to watch video and track the balloon.

4. surviving the journey

So currently, you’re sitting in front of a computer at somewhere around 25°C and indoors where there’s no wind. Suddenly, you’re at 35,000 ft. It’s -40° and the wind is blowing at 150 mph. Obviously, you’re going to be…. eh, a little uncomfortable, but does your computer still work? What about your keyboard, is it covered in ice? Is your desk cracking? Will your chair still support you? These are the things we need to consider for every piece of equipment from electronics, to batteries, to rope. We will need to be assured that they behave satisfactorily in these conditions. So how do we test these things at -40° here, on the ground, in Kentucky?

We built a cryo-chamber, of course. We started with an old chest freezer, wrapped it in layers of home attic insulation and then added insulation on top to form a lid. 2 chambers were built inside. On the right is the test chamber. On the left is the dry ice chamber. Computerized fans control transfer of the cold air from the dry ice chamber to the testing chamber. Temperature sensors are placed inside on the item to be tested. A microcontroller controls the fans and reports the temperature via a web page. We place 7-10 lbs. of dry ice, which is -78°C (-109°F) when testing. After cooling to the desired temp (either rapidly, or slowly, as needed, the items are taken out and exposed to the stresses they are expected to encounter during the mission. (OK, smacking a -40° bag of alcohol against a brick wall is much more than we would expect during flight but it was interesting). If the material survives, YAY! If not, another material or solution is explored.

Here’s a recent test of a valve design made from solvent welded acrylic.

5. death from above, bring in the lawyers.

So what if this thing crashed down on your parked car? You’d probably not be happy with us. Worse yet, if it crashed on you or got sucked into the engine of a commercial airliner? After all, we are flying at the same altitude as commercial airlines for exactly the same reason.

There are federal regulations that govern these things with all of those concerns in mind. We actually follow the stricter regulations for balloons of sizes much larger and heavier than ours.

Twice each day, 800 weather balloons are launched around the world and none have ever collided with an aircraft. Also consider that there are millions of large birds, such as geese, that navigate the airways without any consideration of rules or safety, and still the chance of a loss of life due to a bird strike is about 1 in a billion flying hours. Thus, the odds of our balloon causing any danger are extremely slim. Nevertheless, we aim to be responsible in this regard and conduct our experiments well above the minimum safety requirements. We don’t want to hurt any people or property and don’t want to reflect badly on amateur scientists.

If our balloon fails or is instructed to land, it is designed to make a soft landing. For one thing, the heaviest part of the balloon, the payload, is encased in styrofoam. It’s cushioned enough and not heavy enough to cause damage to people, animals, vehicles, or homes/buildings. Secondly, if our balloon fails or is instructed to cutdown, it has an innovative parachute.

Upon cutdown (failure or landing), a rope placed in contact with nichrome wire (the wire in toasters and hairdryers) will be heated and the rope will break. This will allow the balloon envelope to invert itself and dump all its helium. The envelope will then act as a drag chute slowing the descent of the balloon and payload. And, yeah, just in case, in the highly unlikely event of property or personal damage, we’re insured, which is one of the significant costs of our project.

The “take-home” messages here are that safety has been thoroughly considered, dangers have been tremendously minimized beyond federal requirements.

Though we exceed federal safety requirements, there are still legal concerns. Foremost is an ordinance in Louisville that prohibits the release of latex or mylar balloons. Though this ordinance is intended as an environmental protection measure to avoid litter and damage to wildlife that may ingest these balloons, it, unfortunately includes balloons used for scientific purposes. Thus, we must travel 70 miles north to Spaceport, Indiana (Columbus, IN) to launch. Since we conduct many test flights, this is a major challenge. We are (and encourage you to) petition Louisville Metro government to make an exception for scientific or educational balloon launches.

One of our recent test flights launched from Spaceport, Indiana landed in family’s backyard in Miamisburg, OH. The family made a video. Check it out here:

6. watching our weight.

Obviously, when lifting a balloon, the weight it’s carrying is a major concern. Every piece of the balloon is evaluated to limit weight and these evaluations are too numerous to mention. Here’s a couple examples though.

The valve that releases helium and prevents air from entering went through lots of revisions. Machining metal was considered, 3-D printing, etc. Many had the problem of creating excess weight. Jon introduced the idea of using Line-X material which is typically used as a truck bed liner as a means to strengthen the structure of something light like styrofoam. Shaping the styrofoam became the next problem. Gary came up with a different solution involving folding paper. The paper structure was reinforced with Line-X and the spring assembly followed. Watch this video to see Gary’s initial presentation of his solution during a meeting.

While we currently use a 3 liter soda bottle to contain the ballast, other weight-saving options are still being explored.

From these examples, we hope you can imagine all the considerations and testing we’ve gone through to achieve our current weight.