This past weekend, we found a rare quiet 20 minutes in the hackerspace, before the day’s Mission Control simulation started. I took the opportunity to throw together a little video peek for you all, to see how we’re doing. I had no time for retakes, so it’s dirty, but you will get the peek you’ve been waiting for. It’s 10 minutes long, but as Brad said, “I think we’ve done at least ten minutes worth of work in the last month or two!” Take a look. (Read on after the vid)
With such a complicated system, the only we know that everything works together is to test it, all together. Gary made a GPS simulator to act like a real gps to our flight computer, but actually replay a predefined balloon flight. Since the balloon robot makes all it’s decisions based on the GPS and our commands from Mission Control, we can make the balloon perform like it’s really flying.
So for the last two weeks we’ve been doing just that. It’s been a lot of work, and every night we run a full simulated flight, and every night we discover another problem somewhere. They’re all being fixed, but we’re running out of time. Luckily (or unluckily) the jet stream winds are not very good for the next few days, so we’re not missing anything.
SB-1 will not be launching this weekend. It turns out that our most significant competition, Cornell University’s tight-lipped Project Blue Horizon, may not try to cross the Atlantic until sometime in April, removing pressure to beat them to Europe this weekend. Also, the Jet Stream wasn’t that great for this Sunday, and the on-the-ground weather was even worse for launching a balloon.
SpeedBall-1 could have flown just fine this weekend, however, there’s significant parts of the telemetry pipeline, a long convoluted series of data transfers, that haven’t been tested very much. This will give us some much needed time to make sure that a silly mistake doesn’t doom our precious robot pilot to a watery grave. Continue reading →
Hi everyone. We’ve been working constantly at getting this Balloon off the ground. Until now, though, we’ve had no way to directly measure how far “off the ground” we’d get. Fortunately, Brad and Dan spent their Friday night building just such a measurement device. When you’re flying a constant altitude balloon like ours, the amount of helium it holds, combined with the weight of the robot pilot dangling off the balloon, directly determines the altitude that you will fly at. When your goal is to cross the Atlantic using the thin Jet Stream, you need to fly at a pretty precise altitude, or you’ll miss it entirely!
Unfortunately, our balloon manufacturer was unable to give us a precise volumetric measurement of the delivered envelopes. Unwilling to bet our sweat and tears on such a response means developing a method to measure the volume of air going into the balloon.
We came up with a few ideas that wouldn’t work:
Buying dry ice, letting it sublimate inside the balloon, and comparing the before and after weights. (It would take almost 110 lbs of dry ice to generate 1000 cu ft of gas! That’s expensive!)
Buying a canister of CO2, and measuring the pressure in the bottle before and after filling. (Compressed CO2 is as expensive as dry ice, and we don’t have accurate enough pressure gauges)
Filling the balloon with air, measuring the pressure inside the envelope, then adding a small quantity of dry ice, and measuring the pressure increase. (Requires a constant volume inside the balloon, which may rip its seams)
Making a “bicycle pump” mechanism, and adding measured quantities of air at a time (This would take FOREVER!)
Placing a temperature sensor and a heater in the air stream, to measure the heating effect in order to measure the air speed, which can tell you the volume that flowed through a tube over certain amount of time. (Would take too long to calibrate the temp sensor and heater.)
The last idea inspired us, though. If you can measure the speed of the air you’re putting into the balloon, and measure the cross sectional area of the tube through which the air travels, you can measure the volume of air that’s moved through the tube over time! It’s a simple equation:
Air Speed (M/S) * Cross-Sectional Area (M^2) * Time (S) = Volume (M^3)
How can you measure air speed? Well anemometers work OK, but they’re big and bulky and would block our tube. Airplanes measure their airspeed using a device called a “Pitot Tube.” It looks like the image to the right.
The Pitot Tube works by measuring the difference between the ambient air pressure and the ram-air pressure. Air is forced into the opening of the pitot tube at high velocity, and the difference in pressure gives you the fluid velocity by the following relationship:
Where (Pt – Ps) is the differential pressure, and Rho is the fluid density.
That seems pretty simple, doesn’t it? Well, we were able to pull it off with parts lying around the LVL1 hackerspace in one evening!
We happened to have a long plastic conical nozzle, about 3 inches long, sold at auto parts stores as part of a miscellaneous kit of vacuum hose adapters. Looked pretty much like a pitot tube already! It also happened to allow some tubing we had ( 1/6″ ID 1/8″OD) vinyl tubing to fit up right out the tip perfectly.
We hot glued the tubing into the nozzle. and mounted it on a thin sheet of aluminum with a cable tie and hot glue. This would allow us to position the pitot tube in the center of the airflow in the filling tube, with minimal disruption in the airflow itself. Disrupting the airflow too much is BAD for pitot tube accuracy. Streamline it, and the things around and behind it. Make sure it points directly into the airstream.
For measuring balloon volume, we don’t need helium, plain old air is fine. So, we called upon our in-house air source: the high-power vacuum cleaner blower head from a shop vac. It would make quick work of inflating the giant balloon.
We embedded the aluminum strip, with pitot tube, parallel to the airflow in the hard plastic extension tube of our shop vac. A simple slot was cut on each side using a dremel cutting wheel, and secured with hot glue. (Use safety glasses!)
The pitot gives us ram air pressure, which is only half of the answer – the other half comes from the static source. Where do you measure that from? Well, we think that it should be measured in the balloon, a ways away from the fast-rushing airstream coming in. To do that, we the open end of another tube of the same type about 4 feet into the balloon, away from the direction we’d be aiming the blowing the air in.
Now, with two tubes to give us air pressure from two different places, we just needed a differential pressure sensor to measure difference at the static location and the the pitot tube in the airstream. Fortunately for us, a few months ago we had designed and built a sensor module, complete with a differential pressure sensor that perfectly covered the pressure range we needed to measure today. We just pressed the static tube and the ram air tube onto the two ports of the sensor, and were ready to measure the pressure!
We connected the sensor board up to an Arduino microcontroller, and wrote a quick program to read the pressure in kPa every half-second. We connected the Arduino to a PC running a serial port data logger to save the numbers as a text file.
The procedure to get volume would be: record the pressure every half second, inflate the balloon to full, paste the recorded pressure list data into a spreadsheet to convert pressure to airspeed, and then each airspeed * time to get volume of air each time period, and integrate that M^3 volume over time. For our balloon, we ended up with 28.04 M^3, which is just about 990 cubic feet. That’s a reasonable answer. We think it’s probably right.
One thing we know for sure, is it was easy, cheap and fast. It should work with helium right on the launch pad, which *might* really help amateur ballooning launches, where it’s really hard to tell how much helium you’ve put in the balloon. We’ll publish some more details on how to precisely replicate this in the future.
We’ve been working very hard to keep all the parallel projects rolling in this massive project. We’re doing pretty good so far. The fund raising still needs lots of help – so please drop a few coins in our pledgie —>
The balloons will be ordered in early november, lovingly crafted to meet our desired float altitude and payload weight. We’ll also be sending a string of sensors and a vent valve along to be incorporated during manufacture.
SpeedBall Flight computer v2 PCB is designed and ready to order, we’ve switched the row of module bus connectors to a more affordable style, and added the ability to cut off the power to the the sensors in the balloon.
Everything we want to put on the balloon needs to be tested at the ultra-cold temperatures that they will fly in. That means we’ve been building a cryogenic test chamber, capable of holding -50C for three days. It may need a little more insulation to cut down on the dry ice usage, but it finally broke through the -50C mark last night.
We’ll be cryo testing things like latex and silicone tubing, hinges, valve seals, batteries, and all the electronics.
The live tracking website for White Star is a daunting project in itself. Thankfully, John Hicks and a few memebers of the Kentucky Open Source Society have volunteered to spearhead that mini-project. It will be our primary information display of the balloon’s live status. It should include all the telemetry that’s being sent down, and display some of it on a map, and maybe a few gauges. For LVL1 mission control, it will provide crucial weather data, performance graphs, and an actual command uplink interface to the balloon.
We’ll be doing some detailed articles here about various aspects of the balloon system as time goes on, so look forward to in-depth info, coming soon. Feel free to comment with questions that you’d like answered in-depth on the White Star Blog.