Thursday 9 October 2014

The Summary

To summarize, I have to say that in the end everything worked as was supposed to, and the whole day went just fine. The payload collected temperature and trajectory data throughout the whole flight, and took some very interesting photos and videos on its way up and partially down. Here are some of the photos. The rest is in the Flickr gallery here: TT7 Flickr
 A view on Roznov pod Radhostem shortly after take off.
The balloon slowly rising above the low altitude mist.
A view on the mountain range above the town of Roznov. Here, restraining the northern low clouds from filling up the valley.
Still below the airplane travelling altitude.
Even the Moon peeked into the view.
One more view on Roznov and the mountain range down to the left. Some cumulus clouds forming in the middle and a massive front coming from the Southeast on the right.
The blackness of space is closer and closer.
Still rising higher...
...and higher.
The maximum altitude 34396m.
And the balloon bursts.
 Accelerating to as much as 36.52m/s.
A long way to go.
Pursuing Aegean Airlines' Airbus A321 on its way from Copenhagen to Athens.
And the farewell photo from the dying camera at about 15.5km.


I made a timelapse video from all of the photos from the payload. The images are too discontinuous to offer a decent spectacle, but it can give you an idea about the flight.


For an illustration, here is a video from around 34km in altitude. The remaining 15 videos are in the Flickr gallery.
These are images from the satellite to give an idea about the cloud cover from a different perspective. From the left to the right 11:00, 12:00, 13:00, 14:00 CEST (UTC+02:00).
Coincidentally, one of the predictions I made in the morning had the burst altitude set to 34434m. Almost spot on the actual maximum altitude. Unfortunately, I didn't prepare a projection with the descent speed as low as 2m/s, otherwise it might have shown the landing site almost exactly. The second picture is the actual trajectory. So to summarize, the expected ascent speed was 3.48m/s versus the actual average ascent speed of 4.21m/s. I think telling the intended helium volume when filling the balloon the way I did it, was not very accurate. Thus I overfilled the balloon a little. The expected burst altitude was 31434m versus the actual of 34396m. In this case, I was pleased and a little surprised that despite the greater amount of helium, the balloon had managed to surpass the level by that much. Almost 3 kilometers. On the other hand, the expected descent speed was way, way off. The difference between 5m/s and the actual value of 2m/s made the landing site 21km further to the west. The causation was definitely underestimating the parachute's capabilities.
This is the trajectory displayed in Google Earth. The launch in Roznov pod Radhostem, the burst 34396m above the village of Morkov and the landing in the village of Jindrichov. If anyone wants to examine it more closely, here is the KML file. TT7 kml And here are all the received telemetry sentences in a TXT file. TT7 txt All the data and data from many more HABs are downloadable from the Habitat website.
This is a diagram of the balloon's altitude and vertical speed. The balloon was launched from 420m and later achieved its maximum altitude of 34396m only to fall and land in 425m. The average ascent speed was 4.21m/s with the 6.19m/s maximum and the 2.95m/s minimum. The average descent speed was 5.45m/s with the 36.52m/s maximum and the 1.59m/s minimum. The whole flight took 4 hours and 2 minutes. The ascent lasted for 2 hours and 13 minutes while the descent took staggering 1 hour and 49 minutes as opposed to the expected 45 minutes. The launch time was 10:17, the burst 12:30 and the landing occured at 14:19 CEST.
This diagram shows the temperature measured by the datalogger. The flight started in 20°C and gradually dropped to the minimum of -45.7°C encountered at 11800m. Luckily, the gloomy predictions of the International Standard Atmosphere model promising -56.5°C didn't come true. Contrary, the balloon enjoyed pleasant 1.7°C at its maximum altitude.On its way back, the balloon went through two local minimums. The first at 21803m of -44.4°C followed by the second at 11204m of -41.2°C. After the landing, the balloon lay in a field on direct sunshine measuring a temperature of more than 30°C. Here is the MSS file from the datalogger. TT7 mss The file format requires a special Cometsystem's software for servicing the datalogger. For download here. Datalogger software The file also includes data from an internal probe, but since the datalogger was fairly isolated from the main part of the payload, I don't think it will be very informative about the actual temperature experienced by the electronic parts.
As I mentioned in the previous article, I wasn't the only one receiving the payload's transmission. In fact, I wasn't even the one who received the most sentences. Since I had to travel between different places during the flight, I was really glad that the payload's trajectory was thoroughly mapped by participants from Poland and Slovakia. Out of interest, I used the Habitat's statistics to find out who was the furthest telemetry recipient. It turned out to be somebody in Eastern Poland about 400km from the balloon. That is not bad for a cheap transmitter.
With all the diagrams and photos already posted, I am slowly getting to the last bits of information I am left with. And since both the flight and the blog are almost done, I should summarize how much it all cost. This is a list of all the things I bought (and remembered) during the preparations. The total cost of this project reached 9027 Czech Crowns which is currently about 329 Euros or 419 Dollars. I should mention that I have saved a great amount of money on helium. I was lucky enough to be allowed to use the leftovers of helium in a local company Retigo which uses it during manufacture. Otherwise, the helium in the quantity I needed could have easily added about 4000Kc or 170 Euros/185 Dollars to the total cost.
And that's about all to be said about my high altitude balloon project. In case you have any questions feel free to leave a comment or send me an email.

Tuesday 7 October 2014

The Launch

By the beginning of September, I was finishing all the work on the balloon. I had already begun surveying the wind conditions. because the upcoming fall meant a danger of stronger winds than I would desire. One window of possible launch days began to materialize around September 15-17. This is the development of predicted landing sites for September 15 and 16.
On Sunday the 14th, I decided that the launch day would be on Tuesday the 16th. That turned out to be a good choice, because on Monday I woke up into an impenetrable fog. In Monday evening, I prepared everything I would need the next day and used the Landing Predictor for a set of tomorrow's possible trajectories.
The following morning, I made the last predictions. The landing sites moved a little bit but remained in my desired zone.
The weather outside looked much more suitable than the previous day, so I put everything into a car and set off to the launch site.
This was the launch site. The wind would take the balloon to the left of the two chimneys in the distance. But first, I had to fill the balloon.
It is good to put some underlay underneath the balloon, so it doesn't get pierced by anything sharp. Also, it is good to have a look at the throttle-valve ahead and make sure you can securely attach your filling instrument to it. In my case, I just slipped the hose on and tightened it with a fastener. It is advisable to handle the balloon in latex gloves as well.
On the other end, I secured the balloon to the filling tube with a pair of plastic fasteners and some tape.
Due to the throttle-valve, the filling to the appropriate volume took more than an hour. Something I didn't count with, so instead of the intended 9AM launch, the balloon lifted off shortly after 10AM.
When the helium inside the balloon began to lift the weight and the set hovered, I knew I had the necessary lift. Earlier, I calculated the neck lift to 1396g giving me 456g of free lift. The gross lift of the helium inside the balloon then was 2196g. To seal the balloon, I used another pair of plastic fasteners, a string and a lot of electric tape to hide the sharp edges of the fasteners. After detaching the filling tube, I folded the balloon ending, laced through the cord going to the parachute and secured it with more fasteners, string and tape.
Then it was time to relocate outside.
I set up the receiving 'station' consisting of the antenna, my notebook and the SDR. I used my phone to create a Wi-Fi hotspot.
After that, I turned on the GPS and verified the lock and the reception. It all worked fine. The only problem was caused either by the excruciatingly slow notebook or the terrible internet connection, but I simply couldn't load the Habitat website and verify that my balloon was visible there. Luckily, that was solved by my friend's phone call a little later. At that moment, I reread my checklist because it was time to turn on the camera script and then finish the payload sealing as quickly as possible.
Once done, a couple of poses for the camera followed, and then finally, the months of preparations were out of my hands. The balloon ascended quickly, perhaps too quickly, because the altitude data showed a vertical speed of about 5m/s as opposed to the expected 3.48m/s.
I traced the payload to about 11km of altitude and then set off for a better located lookout.
At the spot, I was unpleasantly surprised by the low clouds surrounding the whole hill. Originally, one could oversee the whole area from this location, but due to the poor visibility, the only connection with the balloon was the radio
I caught up with the balloon shortly after it surpassed 20km in altitude still rising about those 5m/s and with exactly the same trajectory as predicted.
And then it finally came. The telemetry from 30km in altitude printed out on the screen, and the goal was achieved. Now, I was waiting only for the burst. But the balloon passed 31km, then it past the predicted 31484m, then 32km, then 33km, the 34km. At that time, my eyes were fixed on every new telemetry string enjoying every additional meter. And then finally at 34396 meters, the ascent stopped.
As the balloon began its fall, two things happened. Firstly, I realized that the balloon was slowing down too much. I expected the descent speed at ground to be about 5m/s, but with still 15km remaining, it was already below 5m/s. That meant that the landing site would be further to the west. Another unpleasant consequence was that it would take much longer for the balloon to land. The original expectation was 2.5 hour to ascend and 45 minutes long descent. But after an hour, the balloon had still more than 7km to go with the speed then being 3m/s. My notebook, though being one of those intended to function quite long, was slowly running out of battery. I had to start saving power by periodically turning off the reception, and I decided to relocate further to the west.
After I set up the reception apparatus, I found out that I had about 5 minutes of power remaining. To pile on the bad news, despite the less than ideal internet connection, I managed to display the balloon's current position on a map and found out it was much much further than I would like it to be, heading to a fairly forested and hilly area. The short disbelief was then interrupted by the notebook finally giving up leaving me with the last piece of information that said the balloon was still at about 750 meters in altitude and heading  further to to me unknown terrain. With that in mind, I turned back and headed home to recharge the notebook and come up with a plan.
At home, I turned on my PC and found out that many more people were receiving and uploading the balloon's telemetry, so the trajectory was mapped completely. The last telemetry string was received from 713 meters right above a small village called Jindrichov. I was really glad to see long rows of fields on the map when I switched to the satellite view. The last known position is circled in the picture above. The last shown position was added after I arrived at the spot. At that time, I knew that the balloon had about 250 meters yet to fall and that the descent speed was 2m/s. Based on the horizontal speed, I calculated that the balloon should be within 600 meters from the last known position. That area covered only fields and the village, so I was quite confident about finding it. Moreover, after driving through the village in Google Street View, I could see that the road led through a bottom of a valley and the fields were rising up a hill on both sides. That meant that I could easily find a higher ground and possibly see the payload directly. Thus I set off.
After less than an hour, I was driving through the village looking for a side road leading to a higher ground that I had seen on the map. I took a friend on this journey, and as we were climbing the hill, he noticed something orange lying on the opposite side of the village. I stopped the car, took out a pair of binoculars and saw that it really was my payload lying in a field about two hundred meters from the road. Just to make sure and to upload the last coordinate, I took out the antenna and notebook to do the job.
Then I just drove down to the village and stopped as close to the spot as possible. The field was right behind a couple of houses.
On the way for the payload, I scared a herd of sheep that rather relocated higher up the field.
And there was the payload. Everything that left in the morning was still attached. Only the balloon was ruptured and the antenna due to the impact bent.
After opening the payload, I was more then pleased to see that the photos and videos were all on the SD card. Based on the number of them, the camera managed to function the whole tested time, 2 hours and 41 minutes. The transmitter was still transmitting and the datalogger measuring.
All in all a great day and a cool experience. Time to go home.