Colbert Launch, Part 3: Test Launch

Colbert Weather Balloon Test Flight

With the payload prepared the night before (more like the morning of the launch, since we definitely pulled an all-nighter to ensure everything was ready in time), we set out for the launch destination:

Wilton, NH Middle School

Monte Carlo Analysis - 100 simulations for 4 different launch times. Impact zones in Eastern NH and into Maine. Simulations run based on estimated wind data 2 days before actual flight.

Monte Carlo Analysis - 100 simulations for 4 different launch times. Impact zones in Eastern NH and into Maine. Simulations run based on estimated wind data 2 days before actual flight.

Latitude: 42.846
Longitude: -71.746

Flight Predictions

From the planned launch point in Wilton, NH, the estimated landing zone was roughly 90 miles away (2 hour drive). We planned to launch using our Kaymont 1000g balloon, with a maximum burst altitude at 101,700 feet.

We used our Monte Carlo analysis script to run 100 simulations for our planned launch configuration. A Monte Carlo analysis allows us to vary multiple parameters (such as lift, weight, drag, etc.) to get an "impact zone" of landing points, instead of just a single point. This has proven extremely useful; as you can see, due to the winds there are typically elongated estimated landing regions with a very narrow width in uncertainty, but a longer range of uncertainty in the direction of the winds.

Flight Parameters

Below is a table of our predicted values for the launch, as well as the inputs to our Monte Carlo prediction tool of how much we varied parameters to obtain the predictions shown above.

Launch Preparation

Several friends and family members pledged their help on this flight test. To our great fortune, one of the friends brought along his new toy: a quad-copter drone, outfitted with a 4K camera mounted on a gimbal!

The tracking devices and cameras were activated, the payload was sealed, and we were ready to launch. After a slight 5 minute hold to allow a Southwest commercial jet to safely clear overhead, we counted down, and Colbert Flight Test 1 (CFT-1) was released to the skies!

 

Tracking

Shortly after takeoff, we realized that something was not operating correctly with the SPOT GPS tracker. We had tested it previously and received perfect in-flight updates on the latitude and longitude (no altitude data), but as soon as CFT-1 left the ground, it was radio silence.

As expected (since we programmed it to function as such), the cell phone entered airplane mode shortly after detecting the launch, and so all we could do was wait (and hope) that we heard from it again once it touched back down.

Climb rate was a little faster than anticipated, due to a slightly lighter payload than originally planned, so we updated our flight predictions and set out to head for the planned impact destination. A final Monte Carlo analysis was run based on the actual launch parameters and launch time of 9:56AM EST.

 

Recovery

After a quick pit-stop to refuel, our convoy of two cars (all the other helpers from the launch decided they had already driven enough for one day) neared the estimated landing zone. Finally, contact!! A single SMS message with the GPS coordinates. It touched down a few miles away, so we input the new locations into our GPS and set out. Good thing that we got that one message, since we didn’t hear anything else from either the SPOT GPS tracker or the cell phone after that.

A brief walk into the woods, and victory! The craft is on the ground, with the balloon still caught in a short tree. Thankfully, no drone recoveries, tree climbers, or other long retrieval devices would be required to dislodge it from a tree today!

 

Note: there had been Astronaut Ice Cream attached to the outside of the payload when we launched...turns out they swell in the low-pressure atmosphere and exploded off of our payload. Good thing to note for the actual Colbert launch: poke pin-holes in the ice cream!

Note: there had been Astronaut Ice Cream attached to the outside of the payload when we launched...turns out they swell in the low-pressure atmosphere and exploded off of our payload. Good thing to note for the actual Colbert launch: poke pin-holes in the ice cream!

Lessons Learned

Overall, the mission was a giant success. We got some new learning of what not to do and a few things to improve upon before the actual launch, but it proved out all of our systems prior to the actual Colbert launch.

Several key lessons from this flight test would be added to our High Altitude Balloon repertoire, to help ensure a successful actual launch with Colbert several weeks later.

Immediate disassembly of the payload occurred as soon as we got back to the car to find out what happened to both our tracking devices.

Immediate disassembly of the payload occurred as soon as we got back to the car to find out what happened to both our tracking devices.

SPOT GPS tracker

Although there was nothing obviously wrong with the GPS (still turned on when we opened the payload), something caused it to cease transmissions right before takeoff.

  1. It may have entered an unexpected sleep mode, in which case we (hopefully) would have still received a message once every 24 hours that would have allowed us to track it down.
  2. Some GPS devices are very sensitive to direction and must always be pointing at the sky. However, the GPS was still mounted in the correct orientation, and from past testing we have found that the SPOT GPS still provides reliable tracking even when sideways, upside down, etc. (we conducted a very technical test: throw it in the back of our cars in various orientations and see if it tracks over the course of a month, and it did!)
  3. Spinning – we expect part of the issue could also have been the spinning. While we haven’t tested it (yet), the hypothesis is that too high of a frequency of spinning may prevent the SPOT GPS from providing a GPS lock. Nutations (fancy word that scientists like to use for spinning) were present upon takeoff, as also demonstrated to be an issue for our actual Colbert launch (note: we do have a solution for this now. It’s a combination of climb rate, attachment method of the string to the payload, and adding dampeners on the outside of the box - look out for the blog post in the coming weeks)

GoPro Session

With the test mission's success, we could build up the final, actual Colbert launch payload (above - note the added Blue Colbert filled with counter-weight to help with spinning)

With the test mission's success, we could build up the final, actual Colbert launch payload (above - note the added Blue Colbert filled with counter-weight to help with spinning)

We successfully demonstrated that the GoPro session takes FAR superior quality footage than the RioRand camera, which we flew previously on EFT-4 and EFT-5. It appears in this case, due to the quality and auto-contrast settings and general camera construction, you do get what you pay for.

Astronaut Ice Cream

Another important note – as expected, if you send a sealed bag from sea level to the near-vacuum of space, such as the sealed pouches of Astronaut Ice Cream, it will in fact explode! We thought that the strong packaging and construction of the Ice Cream used may withstand a trip to near-space, but 3 ice creams went up and zero came down, meaning they ruptured upon ascent. Either poking a small pin hole or using our “bag in a bag” technique (to be more technically explained in a future post – think of putting a larger, sealed bag around the smaller sealed bag, to provide an “airlock” to take some of the pressure off the smaller bag) would be used to ensure that the ice cream on the actual Colbert flight makes it safely back to Earth.

Again, as mentioned we are limited in some of the video we can show due to our contract with The Late Show with Stephen Colbert, but stay tuned for a future blog post talking about the actual launch day with Colbert!