The Parachute

We've decided a couple things about parachutes.

1) We're going to have one. And..
2) We're going to make our own.

I picked up some orange nylon fabric for the chute. My brother Brian found a web page with instructions on calculating the shape of the pieces for the chute (http://www.cit.griffith.edu.au/~anthony/kites/parafauna/chute_design/ - and we found out people like to put parachutes on stuffed bears apparently) and together we made it in about three hours. It looks great and works perfectly.

We decided to make it about 36" in diameter. I think this may have been a bit large but we'll see how fast it descends with the weight of the final payload attached.

Cutting out the gores (individual sections):
















Finished chute with test payload, using twine for now:





















Testing:

Cutting the Foam

Note: This is a belated post covering some work done a couple of weeks ago.

We decided to try building our payload container out of foam insulation board. This seemed like a decent solution as it would do a good job of retaining heat, was rather light, and would absorb some of the impact.

The idea was to glue several layers together into a big cube. I figured that we could cut sections out of each layer to fit the components snugly instead of having a big cavity for things to be knocked loose in on impact.

In order to cut the foam boards, I constructed a brace holding a section of nichrome wire. When passing electricity through nichrome wire, it heats up just enough to melt its way through the foam. The brace was to be bolted to the end of my workbench in various ways and at different angles to aid in the cutting.

To power the cutter, I simply hooked the ends of the wire up to a car battery charger. You have to use the old manual style as the automatic ones won't give the proper amount of juice. If you can find one that has a switch to go between 2 and 10 amps, even better. Use 10 amps for long, faster (actually still quite slow) cuts, whereas you can use the 2 amp setting to do slow detailed work without having the foam melt and shrink too far away from the heat.

There is one more adjustment that I found to be good. I added a spring at one end of the nichrome wire. The wire can easily break if too much force is applied while hot. For example, if you get impatient and try to cut too fast. The spring allows a little give and you can visually see that you are pushing it too fast. It also acts as a buffer and allows for a more continuously smooth cut with a few unintentional jerky movements. For example, sometimes the foam may stick on something just a little as you are sliding it across the surface.

The only drawback to using a spring is that it does buffer the movement. The spring stretches as you move too quickly, then it doesn't necessarily settle to a stop where you expect it to.



Here I am cutting a section out for the digital camera. I traced around it with a pencil before cutting. After cutting out the basic shape that was traced, I needed to make a few minor adjustments, but the camera ended up fitting quite snugly in the hole.


At this point we have actually tried several iterations of the camera housing. In this picture you see Greg working on the second one.


It took some practice to get the viewing angle right for the camera. Here you can see one of our tests with the camera inside the layer of foam. In it you can see the board on the edges of the image.


I made the cutting frame removable for hand cutting strange angles as well. This was a rather difficult part to carve out. I later came up with a much better way to do it. As it turns out, you can block off your edges with masking tape. The wire does not get hot enough to cut it immediately and so you can simply ride the line of tape. In this case, I ended up masking off the wedge on both sides of the board and it worked infinitely better. When I was done, I had nice strait edges.

Testing payload package performance on impact

After building a preliminary design for holding the payload we assembled it with the camera installed and did some drop tests off the deck and roof. We experimented with putting a wing on the back of the foam container that would angle the camera down toward the ground after the balloon pops to get some different shots on descent.

While testing, we put the camera in video mode and got footage of the payload falling nose first into the ground. The camera survived all tests, while the foam broke in one spot at the nose on the first impact. We may design the nose to break or deform even easier to absorb more impact and hopefully save the electronics inside.

To start/stop the camera we put a hole in the foam so we could use a tool to press the camera's shutter button.

We had a couple hiccups on the embedded camera's video. The camera's record mode got out of sync causing us to record the opposite of what we wanted to record. After that got sorted out we got some good video. However on the last drop the camera was powered off after opening the package and the lens had not retracted. We're unsure of how/why the camera powered off.










Update (10/16/2010 - Kopper):
Here is a video of a few flights from the camera's point of view:

Wrapping It Up

The spot unit will not exactly have a clear view of the sky as is recommended. Our box will be made from 1.5" foam insulation board and wrapped in a space blanket. While I would guess that the sending of coordinate updates would go through the insulation, the space blanket is radar reflecting, so I figured that I'd better test.

I placed the spot in a quick box constructed from the foam and wrapped it up like a shiny Christmas present in a chunk of the space blanket. I then set it on the back porch exactly where I did the other tests to let it do it's tracking. I thought it might be interesting to see if I did get updates, but they are even more loosely grouped than the previous test. That might indicate that it is not able to see as many gps satellites as it was in the clear.

Update: I left the unit out there for over an hour and had not gotten any updates, so I unwrapped it and the gps updates resumed. I left the container out over night and saw that the distance between the furthest two points was approximately 400 ft. That means a search radius of roughly 200 ft at best.

Spot Tracking

The other day, Greg bought the original Spot One GPS Personal Tracker for $60 on ebay. It arrived and today I activated it in order to start doing some testing.

For the first test, I put it in tracking mode and set it on the corner of my back deck. Tracking mode is what you use to record and share your journey (hike, etc) with others in close to real-time. The spot is supposed to send an update to it's coordinates every ten minutes. So the test was to simply leave it in one spot for over an hour and see how close the grouping of updates was. I figured this should give an idea of it's accuracy when we actually go to recover the equipment.

Note: I assume that the google maps image is lined up perfectly with the actual latitude and longitude, which it likely is not.

The red dot is the actual location that the unit sat for the test. I measured from that location to spot #7 and found that it was 136 feet away. If there are not multiple updates by the time we get to the impact site, we may have roughly a 150 foot search radius.

Simulations

We will be launching the balloon in Eastern Washington. Just to get an idea of how far the balloon might travel, I decided to run a sample scenario or two through the balloon trajectory forecast site. I then took the .kml file that is generated and loaded it up in google earth. The result of the simulation shows that it travelled roughly 100 miles:


I was curious how close the impact was to a road, and what kind of terrain it was shown to have landed in. Zooming in closer showed that it was right in the middle of the river, just about the only water around. What are the chances of that? :)




I intend to run this simulation daily for a while to get a better idea of the range and direction we will need to go for retrieval. After having run a few more, the balloon has always traveled East 80-120 miles. It will be interesting to see if that is generally the case.

Quick Overview

The Idea
We want to send a camera up on a weather balloon to the edge of space (roughly 90,000 ft) in order to get some high altitude pictures of the earth.

This is not the first time this project has been done. We first saw it done by a group of students from MIT. The idea was intriguing not only to us, but apparently to many others as well. In the past couple of years there have been many people to attempt to do it. Many of the projects have failed.

The basic concept is to fill a weather balloon with gas (typically helium, but it looks like sometimes hydrogen or natural gas are used). You then attach a camera for the pictures and some kind of tracking device.

The Camera
A camera normally needs to have a custom firmware installed on it and runs a script from it's memory card. The script controls when the camera takes pictures. It looks like most people simply have the camera snap pictures every five seconds or so. We plan to use an old Canon PowerShot with CHDK on it.

Tracking
GPS is usually used to track and attempt to recover the balloon. Some projects use a cell phone, arduino or the like. We have chosen to keep things as simple and "off the shelf" as possible, so we will be using a spot tracking device.

Batteries
In the stratosphere temperatures can be around fifty degrees below zero (Celsius). This causes issues with most batteries, as it slows their chemical reaction which produces electricity. Some groups will substitute lithium battery packs for the typical lithium ion ones that come with most devices. The lithium variety are much more resilient to the low temperatures. What we are going to do is use chemical hand warmers next to the camera and GPS unit.

The Container
Some sort of container is commonly used to hold your gear suspended from the balloon. You can use some kind of box, an ice chest or fabricate something of your own. The container shields your electronics from the elements and insulates them. We have chosen to make ours from layers of rigid insulation board with hollowed out areas for the payload.