This blog covers the day to day progress of water rocket development by the Air Command Water Rockets team. It is also a facility for people to provide feedback and ask questions.

Monday, April 28, 2008

Our First FTC pressure tests

We did our first T-8 FTC pressure test this weekend. Over the last two weeks we made up a 15 mm nozzle and an end cap that fit exactly inside the FTC. The nozzle is a standard 15mm Gardena style nozzle but one that comes with a threaded section. We made an adapter that allowed us to screw in the nozzle. The adapter has two grooves in it, one to hold an O-ring that seals against the FTC and the other allows the tube to be shrunk to stop the nozzle from flying out. We wrap some wire over the shrunk groove to keep the nozzle in place.

The inspiration for this retention setup came from Urie's water rockets. They have good photos of the nozzle and a good description of the technique.

When we tried to shrink the nozzle end by heating it with a blow torch, we forgot the fact that other parts of the FTC were also going to get heated. *doh* We ended up buckling a part of the FTC near the nozzle. Since this was only going to be a pressure test we didn't really care. Mind you when we were sealing up the end cap we filled the FTC with water, and that went a lot better.

The burst pressure was 190 psi (13.1 bar). The nozzle and end cap stayed in place so we were happy about that. This FTC is really thin walled compared to another length of FTC we got from Damo quite a few months back. The next test will be to wrap some of the glass strapping tape around it and see how much more *crossed fingers* it will hold.

You can see on the left edge where the FTC split.

Saturday, April 26, 2008

Flight Computer Q & A

Over the last year we’ve had quite a few questions about our water rocket flight computers so I’ll try to answer some of the more common ones here:
  1. Why do you need a flight computer, wouldn’t a Tomy Timer do the same thing?

    When you only need a one-shot timer, we always recommend a Tomy Timer as they are inexpensive, you can get them anywhere, there is no need for batteries, very simple operation and light-weight. They are also a proven design.

    We wanted to be able to do more than just deploy a single parachute. A PIC-based micro-controller is an easy way to gain more functionality. Electronic timing is typically more accurate and repeatable compared to mechanical systems. You can also set very short timing delays for things like staging, something difficult to do with a Tomy Timer. There are elaborate mechanical systems that have been built that require no electronics that achieve the same thing.

    A flight computer can have: data logging capability, can control multiple actuators and have in-flight data processing and signal conditioning capability. If you wanted to experiment with active stability, or gather engineering data, it is difficult to do with a Tomy timer.

  2. Isn’t it much heavier than a Tomy Timer?

    With our current design when you add the battery, servo and the flight computer it is perhaps 4 or 5 times heavier than a Tomy Timer. However, we use the battery for both the flight computer and the altimeter, so we save weight that way. The same battery could power the camera as well so you could save weight even further. We haven’t shared the power between the cameras and the flight computer yet.

    If you were to use a small lightweight battery such as the 4LR44, a 4.5g micro servo and surface mount components on a small PCB it could weight about double that of a Tomy imer.

    This weight difference translates to perhaps 10-20 feet altitude loss on small rockets and negligible on larger rockets.

  1. Why do you call it a “flight computer” and not a “timer”?

    The terminology distinction is purely internal to our team. That way we differentiate between our simple electronic “timers” usually based around a 555 or 556 timer and the PIC based ones that have software running on them. Once the flight computer starts processing real-time flight data, the distinction will be more obvious.

  2. Does it do more than timing?

    The published versions of the flight computers mostly do just timing. Although through software they also drive the LED display, do switch de-bouncing and generate the correct PWM signals for the RC servo motors.

  3. Are you looking at adding more functionality?

    We have plans on our roadmap to add more functionality, but we are taking it one step at a time, experimenting with what actually works in the field what doesn’t. Take for example the various G-switch designs we’ve been testing. This involves multiple flights which takes time.

  4. Does your current flight computer control your camera and altimeter?

    The published ones and flown to date have not. The altimeter is powered from the same power source as the flight computer. The Z-log altimeter can be set up to start recording 10 seconds after power-on which means when we turn on the computer, power is also supplied to the altimeter and it starts recording. But there is no direct control between the altimeter and flight computer.

    The V1.5 design has a free port left open to allow the altimeter to be connected to the flight computer through a serial connection. The Z-log altimeter outputs altimeter data continuously over its serial port. However, even V1.5 will not initially have it connected.

    The plan is to feed this altimeter data to the flight computer and it will be able to monitor the altitude and deploy parachutes at preset altitudes or when altitude starts decreasing after apogee. The flight computer will always use the timer capability for backup should something go wrong with the altimeter. At the moment we are working to make the timing as reliable and usable as possible before adding more complex functionality.

    It was always our intention to wire the old cameras to the flight computer so that they could be turned on by the computer just before launch since they only had 30 seconds of record time. However, ever since we bought the new FlyCamOne 2 video cameras with their 30 minute record time, the flight computer/camera integration took lower priority. We start the camera separately before we pressurize the rocket.

  5. Do you have designs that you are keeping secret?

    No. We have no reason to. We only publish the designs once we have flown them a number of times. We like to verify the designs for ourselves before making them public, as it is much easier to fix things before publishing than having to make retractions or corrections later. We find it very useful in making the designs public as other rocketeers help suggested ways of improving them.

    We have already been contacted by 2 rocketeers that have built the flight computers based on our published designs, so we want to make sure we have confidence in the design before they are made public.

  6. Isn’t it expensive?

    Not really. The PIC controller costs AUD$2.84, the handful of discreet components around $10, the batteries are about $3 and the cheap 9g RC servos we get for around $6 each. This means with a PCB the whole electronics ends up costing in the order of ~$25. That is about 1/4 of the price of the camera and about 1/5th the cost of the altimeter.

    Of the ones we have crashed we have been able to reuse most of the parts. Really the only things that do brake are the PCBs, the old G-switches and servos. We have now learned to protect the servos better and have had 2 survive direct impacts since the change.

  7. Will they be available for sale?

    There are currently no plans to sell them in any great numbers as there really isn’t a market for them. Most water rocketeers prefer to build rockets out of inexpensive components. Personally I’d rather be flying rockets than handling order paperwork, chasing payments, etc. etc. We will likely offer 5 of the V1.5 for sale privately at cost price. (Contact us if you are interested - see contact page on our main site) The others we will continue to use for our experiments.

  8. How reliable are they?

    So far we are having relatively good success with deploying parachutes and staging 2-stage rockets with them. All together there have been 68 flights with on-board flight computers, of which 5 failed to deploy and 2 successful deploys but tangled parachutes. This means as part of an integrated recovery system they are about 90% reliable.
  1. What will be in the next version?

    V1.5 of the flight computer is the next iteration we are working on. This version has dual servo capability like V1.4, a loud buzzer for status feedback and helping to locate the rocket lost in tall grass or bushes. One of the new capabilities is that all the timing parameters are configurable in the field and stored in the on-board EEPROM to retain them after power is turned off. There are 15 parameters that are configurable from parachute/staging delays, to multiple servo positions, to the lost rocket sound alarm delays. We are having 9 more PCBs manufactured for this particular design as it makes it more compact and lighter.

  1. Future plans?

Eventually we would like to miniaturize it and use all surface mount components and a much smaller PCB. The final weight and size should be similar to the altimeter (~10grams), although realistically this is at least a year or two away.

Adding logging capability will also be a priority in the upcoming months. We have ideas for air speed sensors that could be used to detect apogee, but have no idea how well they will work or what the data will look like. The idea is to use the normal timing for recovery, and the logging capability to capture data over multiple flights. We will do this for each type of sensor so that we can see what processing will be needed before it can be used effectively for apogee detection.

None of these plans for the flight computer are set in stone and are likely to change along the way. We only work on these during spare time and as a result the development is drawn out.

There have been many people who have flown flight computers on water rockets over the years, many of them a lot more advanced and using accelerometers, logging capability, running science experiments etc. The oldest documented reference I have found is back from March 2000.


The following quote reproduced here in full is taken from a long exchange from the WRA2 forum and is included here because apparently we did not credit Bill with the invention of a water rocket flight computer and that we "stole" the idea from him. (See previous paragraph) In his own words:

Team Seneca Post subject: Posted: Tue Apr 15, 2008 12:13 pm

WRA2 Member
Joined: Sun Dec 31, 2006 4:40 pm
Posts: 97
Location: Seneca, N.Y.

It's not hostility. It's just that I never knew you guys would be so impressed with an electronic timer with a fancy name. I've put real computers on my rockets since the summer of 2005. A computer that does something too, not just a timer. I use an accelerometer to measure the flight and deploy. Back then I also used it to send signals to a small camera to take a snapshot at apogee. I'm the first one to put a computer on a water rocket and it was a real computer, not a tomy timer made from silicon.

Bill W.
Team Seneca


Wednesday, April 23, 2008

Polaron V Preview

Almost completed Polaron V. The main stage is close to 11L and uses a 7mm nozzle and jet foaming to produce a long a sustained thrust curve. The boosters are ~3.35L each and use a 13mm nozzle and normal water to get the main stage up to speed.

(Click on the images to enlarge)

The lower photo shows the detail of where the parachutes are stored on the boosters. It is difficult to see the clear strap holding them in place. A wire connected to the main stage releases the strap and the parachutes can fall out. In theory anyway. We are hoping things don't get tangled on release as there will be three wires hanging from the rocket, parachutes popping left right and center and the clear straps are spring loaded so they will also be in amongst the action.


Quick update

We've been progressing this week with the Polaron IV upgrade. The booster parachute deployment mechanisms are now finished on all three boosters.

I made an extra reinforced bottle for the main stage, but after gluing and heat shrinking I noticed that the coupling was sitting at a bit of an angle. This would have resulted in bent rocket, so I tried to straighten it, but instead of improving it, I managed to break the coupling. :( So I threw that bottle away and had to make up another one. The whole process takes about an hour to reinforce the bottle. We are waiting for the glue to cure before we do a full pressure test.

We've reserved Friday for a full pressure test of the new main stage and new boosters as well. If the weather is favourable we'd like to fly it on Saturday.

We also did some experiments this week in creating foam in a bucket by blowing air through a sintered metal filter into the bubble bath solution. The tiny holes help make foam more readily. We want to try generating foam on the pad to see how it compares with it generated in the air. The main observation was that if the air flowed too fast then the bubbles would re-combine into larger ones, but a slower rate created more smaller bubbles.

There have been some really good discussions on the Yahoo Water Rocket forum this week regarding internal temperatures. It may help to explain why we have had some unexpected failures of the bigger boosters under test. I'll cover this in more detail in the next web update.

Wednesday, April 16, 2008

Progress Updates

We have updated our main site with a few more details of what we've been up to in the workshop. The update also includes some great pictures from the last launch day taken by Andrew from NSWRA.

The update is here:


Monday, April 07, 2008

Tornado Couplings

This weekend we finally managed to produce a number of good Tornado couplings. Tornado couplings connect bottles neck to neck. These one's are easy to make out of gardening supplies from the local hardware store.

Some features:
  • They have a 15mm hole
  • Weigh 13 grams
  • Require no glue
  • Have been tested to 130psi, but can most likely hold more.
  • Require no special tools
  • All plastic construction - no metal.
We have been wanting to make these cost effectively for a while now since we plan on using lots of them to join the spliced pairs of bottles. We will give full construction details in future updates on the main site. We want to put them to use first on the Polaron IV boosters to extend their capacity by another bottle each.

We are also currently working on another staging mechanism design that will be hopefully a lot lighter than the one we have been using on our two stage rocket. Since we are still in the early stages of development, we will describe the design later, once it is more finalised. We have lots of testing and prototyping to do still.

Tuesday, April 01, 2008

2 Stage flights

We had a great weekend launching our rockets at the NSWRA launch event. We flew our newly rebuilt 2 stage Acceleron IV rocket up to 525' (160m). We also flew the Polaron IV rocket with drop away boosters to 510'.

The full update with photos and highlights video is available here:

Here is an panorama from around 500' as the rocket pitched over at apogee.
(click on the image to enlarge)