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.

Wednesday, November 26, 2008

Open Day, G-switch, Acceleron V

We've update our main website with the latest development progress on a number of fronts.

I've spent most of last night doing final work on the FTC rocket to get everything well secured. The rocket has also had another pressure test to 180 psi without any visible issues. Phil had a good suggestion to re-string the nosecone so that the parachute is attached to the nosecone and the nosecone is attached to the rocket. In this way more tangles can be avoided. I've also been a bit concerned at the forced on the top of the rocket where the parachute is attached. The main worry is that a high speed deploy could separate the payload section from the rest of the rocket. I've now threaded another kevlar line through the full length of the payload bay that is attached firmly to the pressure chamber end cap and the other end is looped through the main parachute line. Should the parachute rip off the payload bay, it should all remain connected to the rocket.

We are hoping for nice weather this weekend so we can go flying.

Friday, November 21, 2008

Open Day Preparation

We've been flat out this week trying to get things ready for the open day this Saturday. The main project has been a rebuild of the Acceleron booster. This is our biggest booster to date at ~33L capacity. Acceleron III was 24L and Acceleron IV was 18L because it used 1.25L bottles.

The new booster has three 2L bottle spliced pairs tornado coupled together per segment. There are three segments in total. At 130psi the pyro rocket equivalent is an H310. (~250Ns total impulse)

We have simplified the pod at the top of the booster. The deploy mechanism is the same as the previous versions of Acceleron based on a FC, pressure switch and servo, but the parachute is no longer located in the pod. This meant that the pod could be smaller and lighter. The parachute now resides between the bottle necks of one segment so there is no extra drag. The parachute will be deployed using the piano hinge technique we have been using recently. This should mean the parachute opens sooner. Too many times before the parachute deployed late.


Tuesday, November 18, 2008

Batteries and FC V1.6 progress

I'm always on the look out for small yet moderately powerful batteries to power the on board electronics. It's always a compromise between cost, weight, voltage and current capacity. I've been using the Lithium CR123As in pairs to give me 6V but the amount of current they supply is much more than I need and hence are unnecessarily large. 9V batteries are pretty good in price, voltage and current but are comparatively heavy ~45 grams. A stack of button cells generally cannot supply enough current to drive servo motors.

When cost isn't an issue and you are after minimal weight and maximum power density then I expect some of the tiny Li-po batteries are the way to go. But they do need a charger, are expensive and not readily available.

I came across this video a few weeks back, and finally had a go at pulling apart a 9V battery this week.

Judging by the YouTube video comments a number of issues need to be looked out for:
- Only go for Energizer or Duracell alkaline batteries
- Don't short anything out while doing so.
- Wear gloves should you spill some acid

I only need 6V to power all the electronics so 4 of the cells are all I need. It also means that getting two 9V batteries I can make up three 6V battery packs. At only a few dollars per battery pack it means they are very affordable, and available everywhere. After removing the casing, connectors, and two cells the weight of the battery drops from 45 grams to 25 grams. In comparison a AAA 6V battery pack weighs ~44 grams, and AA battery pack weighs ~90 grams.

I just soldered wires to the contacts to make up the connections. I have started doing this ever since we found that some battery holders under G forces can easily loose contact. They powered the flight computer and servos just fine.

The small form factor means that they can be rearranged to fit the available space.

Flight Computer V1.6

I've ordered all the parts now for the next iteration of the flight computer. This will be a slightly larger run of 20 compared to the 10 made for V1.5. I am waiting for all the components to arrive first in order to verify their dimensions before finalising the PCB. I'm also investigating a small and cheap G-switch option to trigger it rather than the commercial G-switch costing around $9. The PCB will be designed so both variants can be used.

The updated software is also almost finished. We had a request come in from a water rocket team (Oberon WR) who built version 1.5 of the FC but their servo was only turning only around 90 degrees even though the full range of movement was 180 degrees. We tested this with one of our cheap 180 degree servos and sure enough the full range was around 90 degrees. I checked the PWM signal on the CRO and the standard range of ~1ms to ~2ms was correct coming out of the FC. I updated the code to give a range between 0.5ms and 2.5ms and that gave the full range of the servo movement. If I have sent you a copy of the V1.5 code, I can send you the update if you have this issue. Other servos worked fine with the 1-2ms timing.

It seems that some servo manufacturers don't adhere to the standard 1-2ms for full range. This change will also be rolled into V1.6. It may end up as another configurable parameter: 1-2ms for full range or 0.5 - 2.5ms for full range.


Sunday, November 09, 2008

FTC Parachute Deployment Mechanism

We've posted an update on our main website with details of how the T-8 FTC parachute deployment mechanism works.

Included in the update are photos and a short video of how the mechanism works. We are hoping to test fly it at the next opportunity.


Wednesday, November 05, 2008

Sewing on fins

We decided to sew the fins to the FTC rocket. Hopefully this will provide additional strength for the expected G forces. Okay so it's not your ordinary needle and thread but the idea is the same. The wire threads through the split sleeve and through the fin. The wire does not go through the pressure chamber.

PL premium still provides the majority of the strength, but the wire is there to ensure the fin does not come off should the glue fail while the rocket stretches during pressurisation. The wire also keeps the fin in place while the glue dries.

Once the initial glue dries the wire is trimmed and the twisted part is folded down and a more glue is used to create a fillet on either side of the fin. We are trying a technique to minimise the amount of bubbling from the PL, as PL is not very nice for smooth fillets. After we form the smooth fillet with the PL we place scotch tape over the fillet to minimise the amount of atmospheric moisture entering the glue. Hopefully this will slow down the cure rate and won't produce the big bubbles we normally get. Will have to check tomorrow to see how it went.


Monday, November 03, 2008

NSWRA Open Day

NSWRA is having an open day this month on the 22nd November at Toongabbie East Primary School. We will have a water rocket stand there along with the pyro guys. If you are interested in any form of rocketry why not come along? See the flyer below for details:
We are going to be busy over the next few weeks to get the rockets ready for the open day. We want to bring the Polaron rocket with the drop away boosters, the new bigger Acceleron booster, and the new FTC rocket with us. We'll also bring along some of the other smaller rockets.

We have had a bit of a break from water rocket development this last month due to other family commitments, but are now back and should be updating the main site and blog more frequently.

I've also been continuing work on the next iteration of the flight computer (V1.6). The software has had a bit of a revamp to make it easier for the rocketeer. I'm also in the process of laying out a new smaller double sided PCB for it. We've located a local PCB manufacturer with very reasonable rates for the boards so we will likely have around 20 of them made, most of which will go up for sale assembled as was with V1.5. I've switched from using an old version of Protel for the schematic and PCB layout to Eagle which is much more advanced and certainly has some very neat features. Luckily the PCB manufacturer can take the Eagle files directly which greatly simplifies the process.