Thursday 22 August 2013

3am Balloon Launch 5th May (Floatnik 3) Post Flight Analysis.


I've finally found some time to do some post flight analysis on the may 5th flight.
The balloon flight path (southward) and the cut down polygon.   The red dot marks the maximum altitude, this is where the balloon was cut away.

The parachute used for Floatnik 3 was an overly large spare rocket parachute which we had lying around.  It really added to the drift outside the polygon, which is why we set it so far back from the shore.  We also set the polygon well back from Christchurch Airport to make sure we stayed clear. We have had some difficulty on previous flights reaching our predicted assent rate.  This causes the flight to deviate from the predicted flight path quite a bit. We thought it best to give ourselves some safety margin, even though it was the middle of the night and ATC indicated that there was no scheduled air traffic. 

When we recovered the payload in a paddock near Ashburton, the flashing lights made it much easier to find!   Both of the cut downs fired and successfully cut away their cable ties.  This makes Floatnik 3 as the main reason for the flight was to test the cut downs.  Ideally we would have liked to test the cut downs from a much greater height, but we ran out of gas on the balloon fill so our assent rate was down.  The cut downs were set to trigger at 20km or on the polygon, whichever came first.  Given the balloon assent rate, it would have gone far out to sea before bursting so the cut downs did their job admirably.    

The radio downlink worked flawlessly again, we had good reception over the entire flight.  The recovery team got good signal almost as soon as the balloon had been launched, they were about 60km away.  I'm always really impressed at how well we manage to do with a $3, 10mW radio.

Pubudu did some analysis on the raw radio data and made up some nice graphs.
The graph above shows the balloon assent rate which is pretty variable.  Part of this variability might be due to mountain waves as we had southward flow over the Southern Alps.   This is problematic for our balloon prediction code (we basically use CUSF's balloon prediction code running on our own server)  as the weather data it pulls definitely won't cover mountain waves.  

Our overall assent rate is also lower than we intended.  As mentioned above, on this flight we ran out of gas while filling the balloon.  We need to work on our gas metering, I'll cover our efforts on that from in a later blog post.   We had been hoping to get two flights out of each tank, that doesn't look like its going to be workable, so we might have to have another look for a cheaper gas supplier.
The payload decent graph shows some possible mountain wave oscillations as well. 
Lastly the total flight profile.