Servers by jDrones

High altitude balloon glider, test from low altitude balloon

About time I update this

The HAB Glider project aims to return Radisondes from 100,000 feet by an autonomous glider.

Our glider, Clementine, was finally launched under a balloon and released at low altitude.

Here’s the video:

This video saw the progress from the upgrades of the last glider to the test of a low altitude release.

Total list of changes:

  • Changed the aircraft to run missions by airspeed & the TECS controller
  • Added redundancy to release mechanism
  • Long range video, telemetry and radio control
  • On/off ground killswitch & procedure for the release mechanism to avoid unwanted activations
  • Different gauge & length of nichrome wire to burn release mechanism at lower temperature
  • Aircraft uses RTL instead of mission waypoints

We now have 3 methods of release, they all heat up a nichrome wire and melt a fishing line tying all of the release strings together. My project log 9 goes over the first version & testing of this in more detail.

  1. Autonomous release by RPi

monitors altitude through dronekit. Once at altitude, sets the flight mode to RTL, activates a relay & cuts the release string. This also covers a failure mode where if the balloon were to burst before the release altitude, it would notice this, set the flight mode to RTL and cut the release string still.

  1. Release by switch on remote through RPi

Runs in the same loop as the above dronekit program. This doesnt offer any realistic redundancy and is mainly just used for debugging and pilot flexibility.

  1. Redundant release through second set of nichrome wires and relays

This release mechanism is supposed to cover if the raspberry pi were to stop working. It’s activated by a relay connected directly to the Cubepilot orange, and is activated with the remote. The set of nichrome wire on this is much shorter and a different gauge on this one, meaning that it burns hotter. This can cut through the fishing line AND the release strings as well, and if left on for a couple more seconds will burn through itself and open the circuit.

I also changed the aircraft to glide with the TECS controller prioritizing airspeed. This is a bandaid on the original issue of gliding into the wind while trying to cover ground.

The next engineering milestone is to fly to the top of class E airspace. A lot of legal milestones to hit as well.

This plane in its current state can probably pass that, but I want to see if i can:

  1. Change the TECS controller to work on groundspeed, or a hybrid of airspeed and groundspeed.

The idea is to prioritize movement relative to the ground (returning to home while flying into the wind) vs just trying to have a consistent airflow of X m/s over the wings.

  1. redesign the airframe

this airframe was selected and used simply for controls validation. thick wings are comfy in the air but arent the best at gliding

  1. research more into what the optimal way of trying to glide to cover distance is

talking out of my ass here but i thought it’d be fun to see if the brachistrone curve could be adapted to gliders at all:

brachistochrone

probably not since this curve is optimized for speed vs total distance, and drag increases with square of velocity. fun thought though.

1 Like

@junwoo091400 here’s the update to the project :slight_smile:

You made my day!

Loved your style, the humor, the glorius ingenuity for material picking i.e. you new were to put an 1$ model paper and were to put a decent AP unit. Also fully respected the fact you’ve shown your faults and you persistence to success.

Nice story and project, keep it coming!!

Best,
Dimitris

1 Like

We love to see it but when are we seeing it attached to an L3 rocket

supply the L3 rocket u won’t

Glue enough C motors together and you got enough impulse to be L3. Consider it done.

1 Like

Hold best L/D angle of attack. Job done!

BTW, the higher you go, the lower your Reynolds number, the more limited your max AoA will be. This also includes flap/eron deflections too which can give your flight control PID loop some palpitations!

1 Like

oh yeah, idk why that never occurred to me.

No idea how accurate AoA estimation is on arduplane though, it’s probably going to end up being a bit more experimental than it sounds right now.

Servers by jDrones