Monday, November 14, 2005

Actual Plan

I have written a proposal which I have given to some of my potential sponsors. Below is a copy of the relevant parts. I have also named the Balloon The XBR Alpha, for Experimental Robotic Balloon Alpha.



Goal-

The goal of this project is to design, build, and test an automated hot air balloon. The balloon will be able to take off, fly, and land on its own. It will also be able to chart its own path, and eventually, predict landing locations. A laptop computer on the ground will be able to track the balloon using a program created in LabView.


Construction-

The balloon envelope will be constructed from 3/4 oz rip stop nylon, using 12 vertical gores, and one horizontal panel. The fabric is cut into gore shapes, then sewn together to form the entire envelope. The top of the balloon is attached with Velcro, so it may be removed to deflate the balloon. Load tapes, cords running down the balloon to lessen pressure on the fabric, are added. The basket is attached to these cords. The basket is constructed from aluminum, and carries all the controls for the balloon. Normally, the basket carries propane fuel, and a burner. The burner operates by first vaporizing the liquid propane, and burning the vapor to create an intense flame. In addition to these devices, the XRB Alpha will also carry electrical equipment to control the balloon. The heart of the electronic controls is a BASIC STAMP, a small, low power microcontroller that controls the other elements of the balloon. To start, the STAMP will control the burner via a binary servo valve, located in the basket. The STAMP may also eventually control the deflation valve. In addition, the STAMP will also gather information from an array of sensors, including a Global Positioning System (GPS) receiver, an anemometer, ultrasonic altimeter, ground contact sensor, and flame sensor. The role of these sensors is explained under “Flight”. The balloon will also be equipped with a two way serial radio for communication with a laptop on the ground.

Flight-

The balloon will be inflated with cold air by hand while it lies on the ground. After it has been inflated, an initial, human controlled firing of the burner (or burn) will take place. This will heat the air inside the balloon so that it rises to an upright position. Control is now switched over to XRB Alpha by turning on the control systems. The BASIC STAMP will initiate all of the required sensors, and confirm that the ground contact sensor is tripped. The XBR Alpha will then initiate a burn, until data from the GPS indicates that the balloon is rising. Following a preprogrammed flight plan, the

XRB Alpha will achieve a desired altitude, as measured by the GPS, and then begin its descent. To descend, the XRB Alpha will execute fewer burns, allowing the temperature in the envelope to fall, and in turn, allowing the balloon to descend. The BASIC STAMP will ensure that the balloon does not fall too quickly by taking readings from the GPS. The balloon will rely on the altitude given by the GPS until the ultrasonic altimeter detects that the balloon is 9 feet above the ground. The ultrasonic altimeter is necessary because the accuracy of the GPS is not accurate enough to determine the balloons altitude within this distance. Once the BASIC STAMP begins taking readings from the ultrasonic altimeter, it will slow the descent until it detects the ground via the ground contact sensor.

Laptop-

Throughout the entire flight, the XRB Alpha will be in contact with a laptop computer on the ground via a wireless serial radio. Through a program created in LabView, the user will be able to monitor the altitude, location, horizontal and vertical speeds, wind speed, and eventually, a video feed. The laptop will also provide a way to control the balloon manually, in the event that it is necessary.

5 Comments:

At 4:12 AM, Anonymous Angus said...

Hi,

Very interesting stuff.

Do you fly ful size balloons?

I am i pilot of full size balloons in Australia but i have also built my own model using plans i made.

I know of someone in Australia who built a hot air balloon (full size) autopilot, which, basically 'learnt' by being flown on manual for an amount of hours, one it had all this data, it was able to be more acurte, so the more you make it learn, and even learn from its mistakes, the better hit rate of accuracy you have.

I have a few suggestions to what you have done so far, from my experience.

Firstly, what plans are you using to create the full size model? I have plans for a 24 / 12 gore 83cubic meter meter / 2900cuft model with a 35kg lift, which is what i made and used to make my model, i do not have any photos on the net for you too look at but drop me and email at angusp -at- gmail -dot- com.

I assume your basket will take a few hard htis, my suggestion here is a biggie, the first balloon used a wikker basket, the ones these days do too, tried and tested for strength and durability.

When i stiched my model, i double stiched, which was, on a single needel home machine, i sewed one length, then went over it again, creating a second stich for strength.

I have a great source for fabric, so again, just send me an email.

I do not have a vent in the top of my model, as do hardly any other models, because of the tiny area to be heated these thing cool and heat like melting butter and freezing ice water. To let all the air out after landing, i simply unhook the basket then do a 'vertical dump' which is just holding onto the crown line and letting the envelope shoot up letting the air out.

For the electronics, do a ssearch for people who have let weather balloons with a payload of gps/linuxpc/cameras etc up in the air, they have some great info on tracking and the basic electronic side of things.

Like someone else commented, have more than one verticle gore, i have 4 verticle gores.

I dont have loads tapes, simply because they are major overkill,the facric (well at leat what i use) is extreamly strong enough to carry the whole 35 max kg's distributed over it. For attaching it, dont use cord, been there burnt them :) use some cable from a fiching shop, its simple you just cut with pliers then crimp it, i have 12 attachment points sharing the 35kg max load, so at most they are all carrying 2.9kg, i do not carry 35kg's though. My rig weighs hardly anything <5kg so i have a slab of concrete in the bottom to give it a bit more stability and control.

With tanks, i used takes i bought, they are great and if you had say 6 of them, you could fly for an hour with some to have in reserve, i bought my burners, twin burners, off the same people as all my other bottom end stuff.

For valves, dont use a servo puching valve, this requires alot of power and chew's you batteries, i use elecctronic soleniod valves, again from the same people.

I know a bit about electronics eg STAMP's and PICAXE's etc.
I think you should have a look into some sort of WRAP board that will run a small distro of linux, this depends on how much you know linux though, thats what most of the remote weather balloon projects use.

Again, i strongly reccommend some sort of learning system, as this is a proven way to get a higher ratio of sucess.

I have alot more advise and i will be following this project closeley, but i think thats enough for now, i will post some more adive/info later.

Happy Landings :)
Angus

 
At 4:19 AM, Anonymous Angus said...

hi, me again

just found the sites i was thinking of
http://vpizza.org/%7Ejmeehan/balloon/
awesome site, should be lots of help

http://members.shaw.ca/sonde/
interesting too

have fun ;)

ps. im also the owner of the model balloon group ;)

 
At 10:19 PM, Blogger GPS Systems said...

Hi, Thanks for your interesting blog. Keep up the great work! I also have a site & blog about handheld gps, please feel free to visit.

 
At 1:25 AM, Anonymous Darko Novakovic said...

What a blog !!
Thanks man for your efforts

I am considering to make something similar

 
At 5:25 PM, Blogger Paul said...

Hi, I built a balloon very similar to this and now need some plans for building a burner based on two propane bottles as you have. I've found plenty of pictures, but no plans. Let me know if you can share detailed pictures of a burner along with list of fittings, tubing, etc. Thanks

 

Post a Comment

<< Home