2021 - Project Altitude 2.0

Mission:Exceed an altitude of 120,000 feet
Date:Sept 4, 2021
Status:Balloon recovered but failed mission
Altitude:31,659.27 meters (103,869 feet)
Site:Stevens Pass Washington
Duration:

This was a mixed success project. We were able to successfull launch and recover the balloon. However there were a few issues.

  1. The ballon was underfilled so it was only rising about 2 meters per second rather than the target of 3. This ultimately caused the balloon to find equilibrium around 103,500 feet. We assumed the balloon was a "floater" and would be lost as it would not rise enough to burst before the batteries ran out and we would lose track of it. Fortunately it did break and we were able to recover. It is unknown why it broke when it did.
  2. The script on the raspberry pi did not activate properly and thus we did not log any data or get any photos or videos.
  3. The tracker reset during descent and went back into pre-launch startup mode.
  4. The flight stage did not detect descent and move into the high-descent stage. Perhaps it would have if the power had not reset.

Flight Path

Flight Path

Here was our original planned flight path assuming 3.0 meters per second lift. We ended up with only a bit over 2 meters per second which caused it to stall.

Note: temperature is from within the styrofoam payload package

Lessons:

  1. Find more accurate method of getting correct amount of gas in the balloon. Perhaps a simple scale that can hold the weight of the gas canister and fill until it drops to a target. This would ensure that we get the correct weight of gas into the balloon. Using a thrust scale or flow meter is not reliable.

  2. The software and hardware setup needs more error recovery workflows.

  • Add power reset recovery to detect flight stage. If altitude is above a threshold and rising or falling then auto-set the stage. This will allow the pi scripts that receive the stage change events to respond properly.
  • Debug why the tracker did not switch to the high_descent stage.
  1. Need to find a better way to manage the pi in the field. It was set to the home wifi network but could be accessed in the field to configure or update the software.
  • It would be ideal if we could have it communicate status during pre-launch to ensure things are running properly before launch.
  • Perhaps if it was on phone wifi we could ssh into it or send status signals to a web hook during pre-launch stage.

Although we didn't get any pictures from the air, here are a few from the ground.

This is John, the security guard at Stevens Pass giving us the thumbs up that we weren't up to no good in the parking lot.

The balloon parachuted forever. At least it was kind enough to land only a few hundred feet off the road.