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.

Tuesday, November 27, 2007

Polaron IV boosters

We have been working on Polaron IV and its launcher in the background while doing the foam experiments. We have further refined plans now on how the boosters will be attached and how they will separate. The photo above shows three newly machined 13mm aluminium nozzles for the boosters.

Each booster is 90mm wide, 2.3L capacity weighs ~150grams dry including recovery system, has a 13mm nozzle and uses a 450mm launch tube.

The simulator predicts that individually each booster should reach about 130m (430') @ 120psi, however, each of the boosters is intended to lift an additional 900 gram weight. This additional weight being a third of the total weight of the main stage fully loaded with water. Under these conditions the predicted altitude of each booster is around 35m (120'). Giving the main stage a release velocity of ~ 25 m/s.

The calculation gets a little trickier because the main stage also fires at the same time when boosters are launched. Though the main stage only uses a 7mm nozzle and foam so that the overall thrust from the main stage will be much smaller compared to the boosters. The consequence of this additional thrust from the main stage means that each booster has less to lift and therefore will be released at a higher altitude and higher velocity. We will do these additional
calculations when we finish building the main stage and know its parameters in detail.

At 120psi, each booster will produce around 90N of thrust at release and about 190N at the end of the launch tube. That is a combined booster thrust of around 570N as the rocket clears the launch pad. Compare that to the Main stage that will produce around 55N thrust with water only and likely around 35N with foam. From static experiments we found that the main stage should thrust for about 6-7 seconds. Rough estimates for main stage altitude are around 230m (750') at the 120psi level. Actual flight though may vary from this figure if the rocket does not go vertically.

The launcher under construction will have two separate air supply lines allowing us to use different pressures in the boosters compared to the main stage. This should allow us to experiment with a wider variety of rocket configurations. Full plans will be published once the design is finalised and the rocket is tested.

Wednesday, November 21, 2007

Foam vs. Water-only flight test results

We have updated our main site with the results of the foam vs. water-only test flights.

In the analysis we show that foam flights were about 2-3% lower in altitude than the water only rockets. This was a little dissapointing but also encouraging in other ways and certainly gives us a direction for further research with foam. Since the update We tweaked the simulation's drag coefficient and nozzle loss factor so the simulation matched the highest observed water-only flight altitude. We then added the weight and drag of the attached camera, upped the pressure to 120psi and ran the simulation again. The predicted water-only altitude was 350'. The last camera mission on the day was a foam flight with the camera and the altimeter. The altitude measured was 353'. This would put foam on par with water-only.

The altitudes are too close together to be able to make any definitive conclusions one way or another, and a lot more flights are required.

There are a few reasons we want to pursue foam experiments further:

  1. It's a lot of fun.

  2. The residual foam weight issue, described in the update, could point towards a measurable advantage when solved.

  3. These tests were only carried out at low pressures, enough to get the rocket off the ground. As the pressures increase, the take-off and peak velocities will also increase. Due to the difference in velocities between water and foam powered rockets (water = faster & shorter burn, foam = slower & longer burn) the difference in drag will play a more signigficant role since drag is proportional to the square of the velocity. This should favour foam at higher pressures.

  4. Convergent/Divergent (DeLaval) nozzles are yet to be fully analyzed. Although initial tests showed them to be no better at low pressures, higher pressures and nozzle shape optimization are still yet to be tested.

  5. Use of foam may be more optimal for upper stages of a rocket than for the main stage. More simulation is needed.

  6. Efficient generation of foam. From foam thrust measurements we found that foam generated using the Jet Foaming technique produced about 14% less total impulse. We have yet to test alternative ways of generating foam that may be more efficient. A couple of new foam generation ideas are already on the drawing board.

  7. Foam optimisation. So far we have only been using the same ratio of bubble bath to water when mixed, but other combinations will need to be tested.

  8. Different foaming agents. So far we have only tested kids bubble bath to generate foam. There are much better foaming agents available and foam density and viscosity are likely to play key roles in the efficiency.

  9. The test results from all these experiments and data from other rocketeers may be used to build a foam simulation model for further research.
If all the above issues can be optimised for a particular application, then it still may turn out that foam can have higher performance in particular situations.

Monday, November 19, 2007

Great Flight Day

We had an excellent flight day on Sunday. The weather was ideal and none of the rockets crashed. We tested the water-only vs foam flights and got altimeter data for all of them. Virtually all the flights were vertical.

I am in the process of doing a full web update but it will take a day or two to collate all the data and get the videos updated.

The update will also cover bottle reinforcing techniques and their results.


Tuesday, November 13, 2007

Flight Computers & Misc

Most of this week has been spent working on a new design for the flight computer. As mentioned in the web update a couple of weeks back, we want to make it more automated to enable remote operation. We are building our own IR remote so that we can have full control over its operation. Most of that hardware is now designed and part of the remote's software has been written.

The new flight computer will not have any buttons except for the power switch, instead it will have the IR sensor module. It will also have a small speaker to acknowledge mode changes as well as serving as a recovery aid should the rocket fall in tall grass. The flight computer will also support two servos for staging and recovery as well as the capability to support the Zlog altimeters. The flight computer will be able to start the altimeter recording once the flight computer is armed.

Most of the design for this new hardware is now complete, with about half of the software still to be written.

We have now repaired J4III's body and fins, and are in the process of remounting components in the payload bay in order to protect them better during impact. J4III will now also sport the new shock absorbing nosecone to help provide even more protection should the parachute fail.

We have also made our first jacketed bottle that includes a Robinson coupling. We have pressure tested it to 100psi to check for leaks around the coupling, but will need to burst test it to see how much it can actually hold. If the burst tests are successful we would be aiming for launch pressures of around 180psi. While we believe the bottle burst pressure will be around 220psi+ the Robinson coupling is the biggest unknown.

If the weather holds up this weekend we will try to go back and do the foam and water comparison flights.

Tuesday, November 06, 2007


This week we continued working on the stager mechanism. We finished enough of it to test it staging just in the back yard. We used only air at 20psi a couple of times and at 50psi once. Although it worked it was a little stiff to release and that is something we need to look at. We will probably try softer o-rings first. We still have a major design issue to resolve, but have some ideas. (More on this later)

The camera that wasn't behaving during the last launch day appears to be OK, and the cause looks like a poor battery contact. The contacts must have become more compressed inside the camera. We tried it with another battery that has taller solder blobs on the contacts and that worked fine, even when shaking the camera.

We also had a go at building the first prototype of the shock absorbing nosecone. It basically consists of the top 3/4 of a PET bottle which is filled with soft foam. The sides of the bottle have 8 longitudinal cuts all the way around that allow the bottle to easily split and slide over the existing nosecone. It is lightly taped over the top to enable the tape to separate easily on impact. During a crash the padded nosecone slides down and the foam compresses against the internal nosecone, but also the whole thing acts as an air piston to dampen the shock even more. The prototype weighs 35 grams so not a very significant weight penalty. For record flights, this can be removed and the existing nosecone underneath can just be used.

We have also been on the lookout for new launch sites around Sydney with great help from the local rocketry community. We looked at one location (George Kendall Reserve) that looks pretty good and is only about half hour from home.