last updated: 2nd September 2022 - Day 221 - Horizon Deployment

Safety First


Site Index



Rocket Gallery


Where To Buy

10 Challenges




Contact Us


Construction - Basic


Ring Fins

Flat Fins



Construction - Advanced

Robinson Coupling

Splicing Bottles #1

Splicing Bottles AS#5

Reinforcing Bottles

Side Deploy #1

Side Deploy #2

Mk3 Staging Mechanism

Multi-stage Parachutes


Construction - Launchers

Gardena Launcher

Clark Cable-tie

Medium Launcher

Cluster Launcher

Launch Abort Valve

Quick Launcher

How It Works

Drop Away Boosters

Katz Stager Mk2.

Katz Stager Mk3.


Dark Shadow Deployment


Recovery Guide


How Much Water?

Flying Higher

Flying Straight

Building a Launcher

Using Scuba Tanks


Video Taping Tips

MD-80 clone

Making Panoramas


Burst Testing





Servo Timer II




V1.3, V1.3.1, V1.3.2


Deploy Timer 1.1

Project Builds

The Shadow

Shadow II


Polaron G2

Dark Shadow

L1ght Shadow

Flight Log Updates

#221 - Horizon Deploy

#215 - Deployable Boom

#205 - Tall Tripod

#204 - Horizon Deploy

#203 - Thunda 2

#202 - Horizon Launcher

#201 - Flour Rockets

#197 - Dark Shadow II

#196 - Coming Soon

#195 - 3D Printed Rocket

#194 - TP Roll Drop

#193 - Coming Soon

#192 - Stager Tests

#191 - Horizon

#190 - Polaron G3

#189 - Casual Flights

#188 - Skittles Part #2

#187 - Skittles Part #1

#186 - Level 1 HPR

#185 - Liquids in Zero-G

#184 - More Axion G6

#183 - Axion G6

#182 - Casual Flights

#181 - Acoustic Apogee 2

#180 - Light Shadow

#179 - Stratologger

#178 - Acoustic Apogee 1

#177 - Reefing Chutes

#176 - 10 Years

#175 - NSWRA Events

#174 - Mullaley Launch

#173 - Oobleck Rocket

#172 - Coming Soon

#171 - Measuring Altitude

#170 - How Much Water?

#169 - Windy

#168 - Casual Flights 2

#167 - Casual Flights

#166 - Dark Shadow II

#165 - Liquid Density 2

#164 - Liquid Density 1

#163 - Channel 7 News

#162 - Axion and Polaron

#161 - Fog and Boom

#1 to #160 (Updates)


Polaron G2 Build Log


This build log is in chronological order so to see the most recent post you need to Jump To The Bottom. You may need to refresh this page to see any latest updates.

The Polaron G2 rocket will be developed in three separate phases. We are targeting initial launch pressures of 250psi (17 bar) and so most of the equipment will need to be re-tested/upgraded for these pressures. 

Phase 1

The first phase will consist of a 15.9L single stage rocket. This will use a bigger nozzle most likely 15 or 16mm. Phase 1 will test both the rocket systems and ground equipment to the higher pressures and loads.  

Phase 2

In the second phase we will add 3 x 10.6L drop away boosters and extend the main stage by another 3.15L for a total of ~19L. The nozzle will be reduced for the main stage to perhaps 9 or even 7mm for a long sustained burn. Jet foaming will be used to power the main stage while the boosters will use larger nozzles and water only.

This phase will include a set of static ground thrust tests to measure the thrust produced by the jet foaming main stage with different nozzles, to determine which one will be used in combination with the boosters.

Together the three boosters will have slightly more capacity than the Acceleron V booster.

Phase 3

Phase 3 will be broken into two sub-phases.

Phase 3a - A small streamlined sustainer will be developed and tested. The capacity and size of the sustainer is still yet to be determined.

Phase 3b - The sustainer will be fitted to the Polaron G2 main stage. The main stage will be reduced in length and capacity to accommodate the sustainer. The main stage nozzle will be increased again and the main stage will retain the 3 drop away boosters.

CAUTION: If you are going to attempt to build rockets such as these, please exercise extreme care when testing and flying them. This rocket uses high pressures that can potentially cause severe injury to yourself and those around you. Always double check your equipment and review safety procedures before every test and flight. See more information on Safety Guidelines.

Build Log

25 September 2010 - Start of project - See Day 96 for more details and photos. We laid out the rocket on the floor to get an idea of the size of the final rocket. It's going to be fairly large and so we'll need to consider how this will be assembled at the launch site. We also made a fin bevelling jig to give plywood fins a nice even edge.

phase 1

phase 2

Phase 3

Bevelling the edge of a fin on
the bevelling jig.

30 October 2010 - We started creating the reinforced spliced quads. These are reinforced with 200gsm glass cloth. We also pressure tested them to 270psi. See Day 97 for details.

Wrapping glass fibers
into grooves

We use West Systems
epoxy as the resin

Spliced quads
ready for assembly.

Phase 3 main stage
assembled with a
representative sustainer on top.

We also made the nosecone plug out of MDF rings and glued it together. We had our first go at fiberglassing using the plug with mold release. This didn't work out too well and we had to cut the nosecone away from the mold.

Rough cut MDF

Machined to size


First nosecone

Filled and sanded

Final nosecone

27 November 2010 - Second attempt at a nosecone, but this time using a swimming cap as the mold release. See Day 98 for details.

Swimming cap wrapped
on plug

Adding layers of glass

Came away cleanly

We also made the deployment mechanism. This was designed to be an inline mechanism with the nose being pushed off. A servo motor releases the nosecone by releasing a couple of rubber bands.

22 January 2011 - Today we tested the nosecone deployment mechanism on a small rocket. See Day 99 for more information about this launch.

We also made the fin can for the G2 from 5mm Corriflute. We added drinking straws to the edges for better aerodynamics.

29 January 2011 - First launch of the Polaron G2 rocket. See Day 100 for launch details including video. Not everything went according to plan and it looks like the nosecone deployment mechanism didn't work out all that well. We'll have to revisit this.

26 February 2011 - Started working on a new side deploy pod to replace the deployment mechanism. See Day 101.



10 April 2011 - Having repaired the nosecone we also built a backup parachute system that was mounted half way up the rocket. This uses a uMAD to detect apogee so it can deploy the parachute anytime the rocket tips over. See Day 103 for more details.

A spectacular CATO failure of the Polaron G2 today that not only damaged the rocket but also the launcher.

parachute pod


T -5 minutes

That will buff out.

22 April 2011 - We did some thermal pressure tests to try to identify why one of the spliced quads would have exploded. See Day 104 for more details about the thermal tests. We also repaired the nosecone and the backup parachute system.


8 May 2011 - Successful launch day for Polaron G2b. The G2b is a spliced-quad shorter than the full G2. The rocket behaved itself and performed great. This also marks the completion of Phase 1 of this project. See Day 105 for the flight day report.

Repaired Polaron G2b being
prepped the night before.

Little Axion next
to big cousin.

The top section of the rocket
is screwed to the lower
half already on the pad.

Launched at 210psi.


Downloading flight data

We need a bigger table.

A photo for the boys
to take to school for
their weekly "news".

30 July 2011 - Two more good Polaron G2b flights. See Day 109 for the full flight report from the day.

The rocket uses
3.8L of water.

Arming electronics.

Launched at 230psi.

Both main and backup
parachutes opened just after apogee.


Launch crew doing
final inspections.

Second launch at 240psi.

Hurray ... no need for
repairs today.

7 August 2011 - We did a full scale test of the Polaron G2 Phase 2 rocket on the test stand. We wanted to get an idea of what the overall thrust will be as well as the burn duration. With the 9mm nozzle the burn times were nice and long. I think it will be quite spectacular to see this long burn in action. See Day 110 for more.

We also made the booster parachutes that should bring them back down safely.

The G2 main stage is
attached to the test stand.

The yellow string stops the rocket
swinging sideways during release.
The black string stops the release
 head from hitting the ground.

9mm nozzle under test.

Finished parachutes.


8 April 2012 - Launching the full size Polaron G2 at 220psi. See Day 118 for full flight report.

Top section is screwed on.

Filling with 4 liters of water.

Launched at 220psi.

Having faced high acceleration
and landing the rocket now
must deal with the eager recovery crew.

14 August 2013 - After a pause in development we have started this project up again. As an interim step we are going to try launching a smaller reinforced rocket first with smaller boosters. The purpose of this test will be to test new booster attachment mechanisms that should be able to hold the boosters at the higher pressures. We are going to be using the Axion G4 rocket which has flown a number of times. If the tests go well we'll apply the same modifications to the Polaron G2 rocket.

20 August 2013 -  In order to continue development of the rocket, we needed to make a couple of fixes to the launcher. Unfortunately a box fell on it a few months ago and damaged one of the nozzle seat mounts. We repaired the launcher today and machined up a stronger mount that was then soldered to the nozzle seat and the sliding plate. We've done this to two of the seats so far, and when the last one breaks we will do it to that one as well.

We also added a new longer lever to the launcher to make it easier to release the rocket at higher pressures. The control panel was also tested and a couple of o-rings replaced that were looking a little worn.

Newly soldered nozzle seat mount

Booster fill tubes fitted to launcher.

Attached a longer release arm.

Testing end to end connections
between panel and launcher.

The intention with the smaller rocket is to launch the rocket at higher pressure (200psi) while still using lower pressure (120psi) for the boosters. Our control panel has the dual pressure capability so we are going to use it. We are also going to fly the main stage with a 7mm nozzle to stretch the thrust out even more. The 7mm nozzle isn't enough to launch the rocket by itself and so the boosters should get it up to reasonable speed. This is more or less a scale test of when the Polaron G2 is launched with a 9mm nozzle.

We already have a brass 7mm nozzle so we are going to use it on the main stage.

22 August 2013 - One of the first things that needs sorting out is how to hold onto the boosters securely especially when using higher pressures and bigger nozzles. The glued on pins and tubes we used on the Polaron series just aren't going to cut it. Each booster will exert a force of approximately 600N (61kgf) on the attachment point. That's a total of 1800N (184kgf) pushing on the central stage. Add a safety margin to it and you are looking at over 2000N (200kgf) That's a couple of larger adults hanging off it!

The fixed tubes had a couple of disadvantages in the design flexibility. So the solution we are testing is a clamp that sits against a thrust ring attached to the outside of the rocket. Some advantages of the clamp approach include:

  • We can detach it and move it between rockets.
  • We can adjust the positions of the individual loops to fine tune where the booster sits
  • We can attach a rail button to it
  • It can clamp the fin-can under it.

We made the clamp today to fit a 90mm fiberglass reinforced airframe. The ring is made from aluminium to reduce weight, but the loops are made from steel to stand up to the concentrated forces without deformation. Each loop was attached with a pair of 1/8" counter sunk screws. We then used a Dremmel tool to cut off the excess thread. The clamp weighs 38 grams. We put a strip of electrical tape on the inside to help protect the fiberglass airframe and provide for better friction.

Fin can is held down by the clamp

Booster clamp

Detailed view

Test fitting clamp with fin can

23 August 2013 - We made the thrust ring today. The thrust ring is made from a pair of fiberglass rings cut out from an old 110mm reinforced splice. The ring is 12mm wide.

Fiberglass thrust ring is made of two layers

24 August 2013 - We attached the thrust ring to the rocket body with 24Hr Epoxy. We used rubber bands to hold the rings closed while the glue cured. In order to simplify the launch procedure we decided to set the pressure regulator at 210psi and we will keep it there. During the fill we will only use the individual line valves to pressurise the rocket and boosters to their specific pressures and shut them off when the target pressure is reached. We have to be careful not to over-pressurise the boosters.

Test fitting the rocket and boosters

Finding the location of where to put
the booster clamp and thrust ring

The rocket was sanded prior to gluing

Thrust ring glued in place

25 August 2013 - Due to the smaller diameter main stage, we didn't have enough room to use our normal guide rail. So we needed to make a narrow guide rail. We bought a 3m long aluminium extrusion from Bunnings that has has a circular profile that looks like it may work. The intension is to use it with our regular legs to hold it upright. Due to the different profile we needed to make a pair of new smaller rail buttons. We cut the rail down to 2.5m so we could easily transport it in the car.

One of the rail buttons was attached to the booster clamp and the other to a flat plate that will be attached to the side of the rocket.

Profile of the new rail

New rail buttons

Rail button attached to the
side of the rocket

Rail button attached to clamp

27 August 2013 - We made 3 short aluminium tubes and glued them to the side of the lower fairing. These guide tubes are not load bearing but slip over the top pins on the boosters. This helps keep the boosters aligned with the axis of the rocket during boost as well as during burnout when the boosters start sliding backwards. With the small loops on the clamp, there is a real chance that a booster could get snagged by the pins if it did not slide out parallel. These guide tubes ensure that both bottom pins are aligned.

We also attached the booster parachute release wires to the top of the fairing. These slip into the booster deployment mechanisms.

Guide tubes glued to the lowest fairing

Booster Parachute release
wires attached to the fairing

Clamp held together with a screw

28 August 2013 - Today we made the guide rail. We had to reinforce the rail with a plank of wood because by itself the rail was quite flexible. We made a couple of brackets that would allow the guide rail to slip onto the quick launcher legs.

New rail on the left,
old rail on the right

Comparison of old and new
guide rail profiles

The piece of wood is used
for stiffening the rail


The bottom bracket that fits onto the
quick launcher

The top bracket for the quick launcher

29 August 2013 - We attached the new narrow guide rail to the launcher. We decided to use the quick launcher legs this time rather than the medium launcher legs, as these provide a support point further up the guide rail making it more stable. It is also a lot easier to put up. The bottom of the quick launcher legs is just held down with a couple of wing nuts.

The guide rail is locked to the legs
with a pin.

Pair of wing nuts hold the launcher
legs to the launcher

Working into the night

The guide rail has to be removable so that we can place it on the launcher after the rocket is mounted. This just makes the whole setup process easier.

Bottom bracket in place

Top bracket in place

Cluster Launcher with
quick launcher legs

New guide rail fitted to
the quick launcher legs

30 August 2013 - We configured the new zLog mod 6 altimeter today to record at 10Hz. We used a single LiPo battery to power it. The altimeter was mounted in the middle fairing to help protect it in the event of a crash. We also mounted the HD #11 camera on the side of the rocket on the fairing.

31 August 2013 - We launched the rocket a couple of times today. The full write up of the event is here: Day 137. The booster clamp worked well so we can go ahead and make a new one for the 110mm diameter body of the Polaron G2.

Getting the Axion G4 ready

Boosters just separating

Pressurised to 120psi and 200psi

17 October 2013 - We bought another 20m of 200gsm plain weave cloth today for $189 as we will need more for reinforcing the spliced quads. We have 5 of the quads spliced together and ready to be reinforced.

20m roll of 200gsm e-glass plain weave

26 October 2013 - We reinforced a couple of the spliced quads today. This now gives us a total of 4 for the boosters. We will make up another 3 (1 spare) and then we should have a full set for construction.

Dad and I have also continued discussions about the launcher. We have been wanting to make a new launcher with all nozzles individually retained because of the loads involved. After considering the complications with ensuring simultaneous release we decided to reinforce the current launcher. The launcher is simple and has been reliable. We will also need to reinforce the main stage nozzle since it is going to be carrying all the load. We are going to make up a new nozzle that will protrude into the bottle neck to help stop it from collapsing and we'll make a new casing that will cover the cap to help support the nozzle. We'll also replace the rubber seal between the bottle and the nozzle with an o-ring.


Getting ready for more fiberglassing

28 October 2013 - It's only 2 days later but we've redesigned the launcher yet again today. The previous solution wasn't quite scalable for bigger rockets or higher pressures. The next launcher we build we want to use for future rockets with more capability and the previous solution would have only been acceptable enough for the G2, We would have to redesign the launcher again later anyway for the new rockets. So we are biting the bullet and we'll build the new launcher for use with the G2. It will also be a good way to test it.

The biggest issue is the exact simultaneous release of all the boosters and the main stage. We want to delay the release of the main stage slightly after the boosters to ensure the boosters are fully engaged when the main stage is released. At higher pressures this becomes even more critical as even small delays can result in the boosters getting left on the pad if the main stage takes off a little sooner.

Another complication is that we want to make the nozzle's radial position adjustable for different sized rockets as before and allow for some movement during pressurisation as the bottles expand. We looked at the typical way simultaneous release is done with low pressure clustered rockets and the one plate joining all the nozzle collars together. However, this does not translate well to high pressures where more force is needed and still allow us to move the nozzles in and out.

We also considered machining large quick release connectors with ball bearings to hold the nozzles, but the machining would have been difficult because of the accuracy required and would have restricted us to a limited range of nozzle sizes. The pull back force on the quick connector collars also would have been large especially with big nozzles and high pressures. We also looked at spring assisted releases for the plate, but that would have taken a lot of force to compress the springs initially.

We finally decided on an electro-mechanical approach to holding down and releasing the nozzles individually. Each nozzle would be released by a separate solenoid. The lever arrangement is similar to a lot of spear guns.

This approach has a number of advantages for our particular requirements:

  • We can use the same nozzle and nozzle seat concept as we have now with the Acceleron cluster launcher.

  • We only need to glue an aluminium ring with 1 or 2mm wall thickness to the outside of the existing nozzles.

  • The lever is normally dropped down so you can put the rocket on the launcher without a problem. You don't have to compress any big springs.

  • Once the nozzle is in place, you lift the lever to lock it and extend the solenoid plunger to keep the lever engaged.

  • When you pressurise the rocket the lever puts a much smaller force on the solenoid. Perhaps only a few kg. Depending on the lever ratio. so you don't need a big solenoid and not a lot of power to drive it.

  • The force of the booster pushing up will move the lever quickly out of the way once released.

  • The whole mechanism is attached to the nozzle seat so it can slide in and out for bigger or smaller diameter bottles easily as there is no mechanical linkage to the other nozzles. You can have as many boosters as you like without a need to change the release mechanism.

  • You can easily move the mechanism relative to the nozzle seat for different size nozzles.

  • Because it is actuated electrically we can synchronize all the nozzles at the same time no matter where they are or how many you have. We can also release the central one a fraction of a second later to the others.

  • If we use higher pressures then we can use several of these mechanisms on each nozzle to share the load. We could have 3 or 4 levers holding down each nozzle.

  • There is no need for very accurate machining, all can be made fairly simply with hand tools.


Electro-mechanical release with solenoid

A small selection of some of the
launcher design sketches over the last few months

29 October 2013 - Added another refinement to the release mechanism design to remove the lateral load on the solenoid plunger. This should also allow for greater loads as the second lever holds all the load instead of the solenoid plunger directly. A small spring will be used to keep this second lever positively engaged into the main lever arm.

Solenoid release with secondary lever

2 November 2013 - Today we bought a couple of aluminium channels to start on the release mechanism prototype. We made the lever and support to test how well it would move. For the pivot we used 2.2mm steel wire. After more consideration we decided to switch this to 3mm stainless steel rod to help spread the load more.

Prototyping lever

6 November 2013 - We cut out more fiberglass for 2 more spliced quads. Unfortunately we ran out of time to do the actual reinforcing which will have to wait till next week.

14 November 2013 - We further refined the release mechanism today to make it more compact. The solenoid was positioned under the first lever and the second lever was pivoted in the middle. This will also help get a further 2:1 lever ratio for the release. This also made it easier to mount the lever and solenoid. We continued to cut out the prototype parts to test the mechanism.

Prototyping mechanism with solenoid

15 November 2013 - Made the loops for the booster clamps. We are using 6 loops on the bottom clamp with 2 loops for each booster. These are the load bearing ones. We are using another 3 loops at the top of the boosters to keep them pointing in the right direction. There are no longitudinal loads on these only lateral. Only one loop is needed for each booster at the top.


16 November 2013 - Dad and I spent a couple of hours fiberglassing another two of the spliced quads. That now makes 6 which is enough for all the boosters. We still have one more to do as a spare backup. It takes us around 40 minutes to do the layup.  We used the GoPro's timelapse function to film the entire process. (see below)

17 November 2013 - We sanded the 4 spliced quads today with the Dremmel tool to remove all the sharp bits sticking out. We'll need to pressure test them next probably up to 250psi. I think we'll be aiming for 210psi for the first launch, and if everything goes well we'll increases the pressure again for the subsequent launches.

Sanding the sharp bits off the segments with a Dremmel tool

Six booster segments ready for pressure testing

Here is a video of the fiberglassing process:

18 November 2013 - We test assembled the rocket today to get an idea of the overall size and what kind of access we'll have to the rocket. We just taped the boosters to the rocket for the photos below.

Just from a rough calculation the overall capacity will be ~56L including the boosters and will use around 16-17L of water. We also ran the volume of air required through the launches per tank calculator and found out that at the eventual target launch pressure of 250psi, we'll be able to only get 2 launches out of a full 10L scuba tank! This means we'll need to think carefully about aborting launches (careful prep) and will need to bring at least a spare tank to the launch.


20 November 2013 - We made the first of the ring clamps for the main stage today. This ring clamp will transfer the entire booster thrust to the main stage. It is made from 1.8 x 14mm aluminium flat bar and weighs 25 grams. With the loops and screws holding them down, it will be around 52 grams in total.

Main stage ring clamp

22 November 2013 - We made the rest of the ring clamps for the boosters. These will have the pins attached to them. We also made a couple of the test pins so we can do some load tests against the loops. We are still not sure how we will implement the top clamps, loops and pins. These don't need to support any thrust but are used to keep the top of the boosters aligned with the rest of the rocket. These mostly undergo small lateral forces.

Main stage and booster ring clamps

Prototype pins to attach to boosters

Pin and loop mated together

12 January 2014 - We replaced the 2.2mm steel wire with 3.2mm stainless steel rod as the pivot for the lever arm. This should help distribute the load better from the lever to the rest of the release head. For this particular launcher there will be over 300N applied to this pivot pin.

New pivot pins

15 January 2014 -We machined the prototype nozzle seat today from a 7/8" brass hex stock. The dimensions are the same as the cluster launcher that we normally use as these will fit the existing nozzles. The main difference is that this nozzle seat has the air inlet from the side rather than from underneath. This just makes it easier to mount and test the prototype.

We also made a small PVC ring that will go over the nozzle and will be what holds the nozzle down.

Machining Nozzle seat

Nozzle seat finished

PVC nozzle retaining ring

It will be glued in place like this

17 January 2014 - We tested the solenoids we have and they didn't supply quite enough pulling force to have enough margin at the higher pressures. So we decided to mount a servo motor on the side of the mechanism so that we can test the release head and whether it can hold down the nozzle at the full pressure. For this test we went with the release configuration described here. If we later decide to get larger solenoids we'll replace the lever and locate the solenoid on the bottom. The servo motor we are using is a MG996R which is quite strong at 10Kg/cm.

Nozzle seat attached to frame

With nozzle in place

Servo motor used for tests

18 January 2014 - We finished making the changes to the release prototype. The whole mechanism now fits together and can be easily mounted. We also glued the PVC ring to the nozzle with the usual 24 hour epoxy.

Here the servo is attached
to the side of the frame

View from the other side

Close-up of how the lever grips
the PVC retaining ring

19 January 2014 - Dad machined 4 new full-bore tornado tubes today. It's quite an involved process and took around 2 hours. We needed an extra 3 tornado tubes for the boosters, as our other full-bore tornado tubes are tied up in other projects.

We also finished making the prototype release head today. The release head was just screwed down to a flat board for the initial tests. We used the Servo Timer II with a remote trigger to activate the servo. We put it and the 9V battery in a plastic bag to protect them from any water during the launch.

Dad changing the gears on the lathe
to allow us to cut the right threads

Dad machining the full-bore
tornado tubes.

Four tornado tubes ready for action

The first test was carried out at 20psi with air only just to see how everything held up. The release was clean with no problems. On the second test we went up to 35psi again with air only. This was also a nice clean release. 

We then set up the old medium launcher and mounted the release mechanism to it. We wanted to test the release head at higher pressures but didn't want the bottle flying off into the neighbours yard so we tied it down. The guide rails were meant to guide the bottle so that it could clear the release head vertically but keep it contained. The first test was carried out at 100psi with the bottle completely filled with water. The test went well again and the release was clean.

The last test was done at 200psi, and again the released head had no trouble holding down the nozzle or releasing it. Post test look at the PVC nozzle ring, and the hold down lever arm showed no sign of wear or bending which is good. We reviewed the slow motion video of the release process, and found that from the time the servo timer gets triggered (LED goes out) and the time the rocket starts moving was around 62ms. This is pretty slow, and we are hoping that if we'll need faster action a solenoid will be able to do better. The speed is important so that all nozzles release simultaneously. Although if they are all delayed by the same amount then that should be ok.


Setting up to test the release
mechanism at full pressure

Here is a short video of the tests:

So good results overall and we are happy to progress with making the rest of the launcher and rocket. The focus is now on getting this rocket and launcher finished for the NSWRA high power launch coming up fairly soon.

27 January 2014 - Lots of work done today:

Bought all kinds of materials for the launcher, including aluminium extrusions, the base plywood, and electronic bits and pieces for the launcher.

We made the remote launch control box with the arm and launch buttons. The arm switch uses the classic rocket switch protector to prevent accidental firing. The controller is very simple in that it only closes a wire loop. The trigger is made in such a way that you need to close the loop to fire. The arm and launch switches are in series. This was done so that an open circuit won't accidentally launch the rocket. We also got 20m of heavy duty wire that will run to the main controller at the launcher. The intention is that there will be 4 Servo Timer IIs to each control a servo. They will all be triggered simultaneously, but have perhaps 30ms delay. I'll update the firmware so that instead of the configurable 1 second delay increments, it will be selectable in 5ms increments so that we get +/-30ms delay range. That way we'll be able to tune each release head timing to get a simultaneous release of all 4. We'll use the high speed camera to film the releases and adjust the release heads as necessary. The main stage will have a slight delay to make sure the boosters are fully engaged.


Remote Launch controller

Booster clamps

Servo mounting brackets and
Secondary levers

The booster and main stage clamps have had their M4 bolts put in. Each bolt holds the clamp in place and prevents the clamp from opening.

We cut out the remaining 3 servo motor mounting brackets and drilled the appropriate mounting holes.

We have decided on the mounting method for the boosters that will allow them to slide radially to fit different size rockets and also allow for bottle expansion during pressurising. On the old launcher we had the slides go through the base, as the air was being fed from underneath. This time we are mounting everything on top of the base.

28 January 2014 - Made up more brackets for the booster release heads. We also made up the sliding brackets for the release heads.


Sliding brackets for the release heads

30 January 2014 - Today we cut the base out of a sheet of plywood. The detachable legs will be mounted underneath, and the launcher will be mounted directly on top. The air for the boosters and main stage will be fed through the top. This makes the arrangement a lot easier and requires less parts. We will use normal scuba hoses and directly connect them to the nozzle seats. Each hose will be connected to a central distribution head. This again simplifies the entire air inlet manifold, and gives us the ability to easily switch from single pressure to dual pressure system when needed.

We also drilled out all the holes for the slide brackets and located them on the base. We wanted to make the fill tubes removable so that the entire launcher can be transported, but in the end we decided to make the whole release head removable so that the launch tube can be permanently attached to the rest of the release head. We will mount these last as we want to carry out synchronization tests with small bottles first.


Cut out base with drilled out holes
in brackets

Sliding brackets assembled and
partly screwed down

31 January 2014 - We attached the release head bases to the sliding brackets.

1 February 2014 - Made the main stage release head framework and started assembling the remaining booster release heads. We also bought 500mm of the 7/8" brass hex stock so we could machine the rest of the nozzle seats. The off cuts we had just weren't long enough. The brass was bought from Edcon Steel for $25.


Release head frames are mostly done

500mm of brass hex bar stock

4 February 2014 - We cut out the main lever arms today and drilled all the pivot holes. We also received two more servo motors so we have enough for the whole launcher.


Rough cut primary lever arms

New servo motors ready to be mounted

5 February 2014 - Dad made up a series of scuba hose adaptors that will be used to connect all the hoses to the launcher. We had a radial distribution manifold that all the hoses connect to that we were going to use for the launcher, but after looking at how the hoses sat, we decided to make a new manifold that would make it easier to locate the hoses.


Scuba hose adaptors

Radial distribution manifold

6 February 2014 - Dad machined up 2 more of the booster nozzle seats today and made the hose manifold as well. We need to run some simulations first to see what nozzle size we will use for the main stage. The original static tests were done with a 9mm nozzle, but we are likely to make that 10mm just to increase the thrust of the main stage a little bit. We need to be careful not to make it too big because the thrust should be lower than the boosters. With the fill tubes acting as launch tubes, the boosters actually produce lower thrust while on the launch tube, than they do after they leave it. The fill tubes are 12mm diameter and the nozzles are almost 16mm. Dad also found a non return valve lying around which already fits the standard scuba threads so that was an easy win. All the connections use standard scuba threads so it makes it very easy to connect, or reconfigure things.

We also made the pivot pins for the lever arms out of the 3mm stainless steel rod.

New booster nozzle seats on left

Linear distribution manifold

Non-return valve for the main stage


Manifold components

Assembled with non-return valve

Booster hoses connected

Pivot pins for release heads with
temporary o-rings to hold them in place

We'll need to drill the hose connectors next in the nozzle seats, but we have to see which way the hoses will be connected first so we can make the holes on the correct side.

8 February 2014 - After much consideration and simulations about what size nozzle to use for the main stage and not coming to any concrete solutions, we decided to make the nozzle size adjustable. The problem is that the nozzle seat must be machined to snugly fit the nozzle diameter, if we wanted to use different nozzle sizes we'd need different size nozzle seats to match. The release lever would also need to be changed if the outside diameter changed. We settled on the same size nozzle on the inside and outside as the boosters so we could use the same nozzle seat dimensions as well as release lever. We then extended the nozzle by around 5mm and made a plastic insert with a smaller hole. This is then press fit into the top of the nozzle. A small ridge on the inside of the nozzle stops the insert from coming out during flight. We can now vary the main stage nozzle size simply be replacing this insert. The nozzle size can be varied from 2-13mm. The insert at the moment has a 9mm hole. Dad machined the new nozzle, insert and nozzle seat today.

Dad did a lot of machining on the day

Here the main stage nozzle is being made



The new main stage nozzle

The plastic insert allows us to change
the nozzle size

9mm insert press fit into the nozzle

We also drilled out and threaded the other nozzle seats to fit the hoses. We had a few goes at laying out the hoses so that they sat neatly. This is the final configuration we settled on: 

Trying out various hose layouts
The central release head is still only
bar stock and no hose are attached

Final hose configuration with a
quick connector for the main air inlet.

Side view with servo motors fitted. The
air manifold is still to be mounted.

Because the non-return valve prevents us from depressurising the main stage remotely, we added a pressure release valve on the body of the main stage nozzle seat.

With the sliding independent release heads we also realized we don't have to arrange them radially, they can be arranged in any other orientation so they can be set up for 2 booster launches, or independently all together and launch drag racing rockets.

Lastly we also decided to replace the plastic retaining ring on the nozzle with an aluminium one that is 0.4mm wider. The plastic ring was always going to be a potential point of failure and so we will machine new ones and glue them in place.

15 February 2014 - Started working on the design of the electronics controller box. We are putting the whole thing in a water proof box with a clear top so that we can see the status LEDs on the servo timers. The power source for the servo timers is a standard 9V battery. For the servo motors we are using a 1.3Ah 12V SLA battery. We expect that each servo could potentially draw about 1 Amp so we are using the bigger battery to power them. Each servo will have its own 5V voltage regulator to deliver the needed current. We could have gone with LiPo batteries, but we already had two of these small SLA batteries so we decided to go with that. There will be a single power switch to turn everything on. We are using a rubber waterproof cover for the switch as well. We are using RC servo motor cable extensions, of which one end will be soldered directly to the board inside the case and the other will have the male adaptor to connect directly to the servo motor. We have opted not to add another set of plugs to the connections to reduce the chance of an open circuit. The remote trigger will be connected via an RCA connector which should be relatively splash proof.

19 February 2014 - We have assembled all the electronics pieces and have started planning the internal layout in the box. With the cover off it will be possible to adjust the timing on the individual timers to synchronize all of them.

Components for the control box.

23 March 2014 We attached leg brackets to the bottom of the base today. We decided to make the legs removable to make it easier for transportation. The legs were also made from wood. The rocket will be held up by the quick launcher guide rail so the launcher won't need to be holding up the entire rocket.

Mounting leg braces


29 March 2014 - Dad machined up the other three nozzle retaining rings today. They will be glued into place once they are fitted to the nozzle caps. Once these are glued we won't be able to take them off the caps. We also made aluminium brackets that will hold the legs to the bottom of the base. After much consideration of how to secure the legs in place we decided on a simple wire pin from the side made from coat hanger wire. The pin is slightly bent which prevents it from falling out.

We also made 6 shims to go under the release mechanism slide mechanism. This allows the release mechanism to slide freely. The shims are 0.6mm thick and made from a plastic folder. We used the same shims under the leg brackets to let the legs slide in easily.

We also attached brackets to hold the air manifold as well the control box to the base. The remaining 3 primary levers for the other release mechanisms were also made today.

Nozzle retaining rings

Attaching shims

Air manifold and control box supports

3 new primary levers


Leg fitted into base

Bent pins to keep them in place

Pin in place

Bottom of launcher

2 April 2014 - We bought new nozzle caps and made holes large enough for the nozzles. We spent quite a bit of time looking for the different caps so that they had a full thread going all the way to the edge. We'll see how these stand up to the pressures. We are using plastic reinforcing rings on the outside of the cap to help keep the thread engaged.

New nozzle caps

Levers fitted to release mechanisms

4 April 2014 - We built the control box today. The 4 individual power supplies for each of the RC servos were built on a veroboard and the servo timers were mounted on the back of it. This allows it to be easily mounted in the box. The batteries, switch and connectors were also mounted in the box. I added the linkages between the servo horns and the secondary levers. These have to be fairly strong, flexible and not stretch so I used some multi-core wire and soldered a loop of it to join the two together. After programming the individual timers start and end positions we tested the whole setup. The tests went well and all 4 release mechanisms looked like they released at the same time. We'll do proper tests under pressure with the high speed camera and only then we'll know if the timing is right.

Completed  control box

Powered up (4x STII)

Launcher ready for testing

Detail of power supply for servo motors

5 April 2014 - It was the Macquarie University astronomy open night. so we brought our rockets and the new launcher for show and tell. We used the demonstrations on the night as a good test to repeatedly go through the launch sequence to see if anything broke or came loose. The launcher also survived transportation to and from the event which is always a bonus.

Testing launcher repeatedly at the
Astronomy open night.

16 April 2014 - We checked and replaced some of the o-rings in the air distribution manifold and hoses. We also tested the non-return valve to make sure the main stage was holding pressure when the boosters were emptied. We then tested the pressure release valve on the main nozzle seat to make sure we can safe the main stage in case of an abort. We tested the launcher in the front yard at 10psi, 20psi and 30psi to make sure everything was sealed and the nozzle release would all happen at the same time.

We then disassembled the entire launcher so we could paint it with water proof acrylic paint.

Testing for leaks

Testing first releases at night

Disassembling for painting

18 April 2014 - We painted the base and legs today with a couple of coats. We also replaced the short hose going to one of the booster nozzles with a more flexible one that allows it to slide easier. The hoses we are using can be cut to length and then the end fittings can be re-connected.

19 April 2014 - We finished painting the third coat today. We also reprogrammed the STIIs to have a shorter arm delay and adjusted the timing to 5ms increments.

3 fresh coats of paint that will hopefully seal the wood

21 April 2014 - With the paint dry we re-assembled launcher. We also added o-rings between the servo horn and the servo body to prevent water from entering them at that point. Thank you to OFlorin for this great suggestion. We coated the o-rings in silicone grease to prevent them from sticking.

We took the launcher to the local oval so we could do higher pressure tests and also look at the synchronization of the releases. We adjusted the timing to get
reasonable results. For one of the tests we delayed the main stage release to see what the timing would be like to make sure the boosters are released before the main stage.

Paul helping set up the high speed camera

GoPro filming at 240fps.

25 April 2014 - After the last set of tests we noticed that 5ms step increments didn't quite give us the fine timing control that we wanted so we reprogrammed the STIIs to give us 2ms field adjustable timing. This gives us a range of 24ms to choose from. We also made up a short removable fill tube so we could test it with some water.

26 April 2014 - We took the launcher to Whalan Reserve today to do three tests. Re-sync the launcher using the new finer control, do one test with water and a fill tube and test each of the release heads at 200psi. All the tests went well and we can now start focusing on the rocket.

Ready for sync test

Adding drag and weight to reinforced bottle

Ready to launch

No worries

200psi air only test

Testing with fill tube and water

20 May 2014 - After some thought we decided to make the fill tubes removable so that we could use different size ones depending on the size of the boosters. So we made up extensions to fit on the nozzle seats with a thread that allows us to screw a fill tube over the top. This also makes it easier to transport the whole launcher. We just epoxied the extensions onto the nozzle seats. Dad then cut threads into the aluminium tubes which make up the fill tubes. We can make easily make up a set of these for the different booster sizes.

Thread adaptors for nozzle seats

Threaded fill tubes that screw onto the adaptors

Thread adaptors fitted to launcher

26 May 2014 - To move to the next stage of testing we decided to launch a small low pressure rocket from the launcher to see how well the synchronization works. We used the same booster retaining ring as we used on the Axion G4 tests. For this we had to attach a thrust ring to the main stage. This was achieved by gluing 2 PET sleeves to the main stage with PL premium. The rest of the rocket was normal as were the old Gluon boosters.

Thrust ring glued to the main stage

Fill tubes on launcher

Boosters fitted on launcher

30 May 2014 - Final prep for the launch. We attached the narrow guide rail to the launcher that allows it to fit between the fins and the boosters. We also further water proofed the launcher by taping up the servo motors and any wire connectors that were exposed to the water streams.

Final prep the night before launch

31 May 2014 - Test launch day. Today we launched the Axion II rocket with Gluon boosters successfully 3 times from the launcher. See Day 146 for a full flight report.


21st September 2014 - Made up the booster clamps today. We are using the smaller 90mm main stage clamp from the previous tests. We test fitted it to the rocket to see if everything remains aligned.

23rd September 2014 - We attached the booster thrust ring today with the 24 hour epoxy. The booster was masked off with electrical tape to give the clamp a clean surface to sit on. We held the thrust ring in with a piece of wire. The thrust ring is made from an old reinforced 2L bottle that we cut up. Each thrust ring has 2 layers of the reinforcing.


24th September 2014 - Test fitting the booster clamps against the thrust rings and doing final alignments. We also made up the top booster clamps with only a single pin. We decided to reuse the fairing from the previous test rocket as it already had the loops attached.


30th September 2014 - We attached the parachute deployment mechanisms to the boosters today. These are only temporary for this test as they will be mounted between the booster segments on the final booster. We attached small sections of PET plastic to the side of the bottle and these were held by glass strapping tape wrapped all the way around as normal tape does not stick to the fiberglass. We attached the sections so we could attach the parachute doors to them with tape. We also attached PET plastic steps underneath the parachutes so they would not slip out during boost.


5th October 2014 - Trip to Mulalley launch site to test boosters with Axion G5. See Day 152 for full details.


19 December 2014 - Tested 3 booster segments to 220psi and cut out the rest of the booster deployment mechanisms.

20 December 2014 - Finished deployment mechanism housings for G2 boosters and attached them to the fairings. Made the top main stage clamp and made fill tube extension tubes.

27 December 2014 - We glued the first of the thrust ring layers to the main pressure chamber.

28 December 2014 - We glued the second of the thrust ring layers to the pressure chamber.

1 January 2015 - Machined up fill tube extensions from brass 3x. Glued extensions into the fill tubes.

17 February 2015 - We attached small rail buttons to the rocket to go on the narrow guide rail. We also attached the parachutes to boosters. Finally we also did a pressure leak test on the booster nozzles as we've had problems with them in the past.

18 February 2015 - We re-attached the fins to the main stage. The fins had to have sections cut out of them to fit over the thrust ring and booster clamp. We also prepared a second set of fins in case a booster again broke a fin off on deployment. We also attached the booster deployment wires to the main stage. These wires are used to deploy the parachutes on the boosters.

12 March 2015 - Flight day. The Polaron G2 - Phase 2 has been successfully flown at Westmar a couple of times. Please see day 157 for a full flight report.


  Back to top     

Copyright © 2006-2021 Air Command Water Rockets

Total page hits since 1 Aug 2006: