Each flight log entry usually represents a launch or test day, and describes the events that took place.
Click on an image to view a larger image, and click the browser's BACK button to return back to the page.

Date:  24th April 2010
Location: Doonside, NSW, Australia
Conditions:  Warm 25C, mostly cloudy 20+km/h wind.
Team Members at Event: PK, GK, Paul K and John K

Fiberglass Rockets

We had a very busy week of preparations in getting two fiberglass rockets ready for launch day. It's actually good having the fixed launch schedule, because it forces you to finish things on time otherwise you have to wait another two weeks until the next launch opportunity hoping the weather will co-operate on the day.

Starting to work on the higher pressure rockets is like starting all over again as a lot of components need new construction techniques and things need to be retested or replaced. We also need to make new tools for making the stronger components.

Testing

During the week the last of the 8 reinforced spliced-pairs finished curing and so we tested them to 270psi (18.6 bar) individually. There are 4 x 90mm spliced-pairs and 4 x 110mm spliced-pairs. Thankfully they all held up well to the pressure without any leaks. Because of the higher pressures we also needed to test a whole slew of our regular components to those pressures. The big unknown were the new Tornado tubes we wanted to use. Thanks to Clifford H. for sending us some PET bottle pre-forms which let us pressure test the couplings to high pressures. We used the pre-form to plug the other end of the coupling and we used our special bottle neck / air hose adapter from the other end. The tornado tubes held 270psi without problem.

We then fully assembled the rocket with the tornado tubes and pressure tested it. One of the the tornado tubes sprung a small leak, and so we tightened it further until it stopped. These are merely plastic-on-plastic seals so a little trickier to seal. We pressure tested it again and all was good.

So we assembled the larger rocket and as we were really tightening up the last coupling it cracked. Major OOOPS! So we inspected the other couplings and found one with lighter stress marks showing where the thread was underneath. Although not cracked it was definitely a concern. The next day dad suggested we swap all of the couplings on both rockets to our old couplings made out of polypropylene. Although the first rocket was already tested to the full operating pressure, the stresses in the over-tightened couplings may become worse over time, and we decided to play it safe. So we had to re-test our old couplings with o-rings to full pressure as well. Thankfully they also held up.

We were going to use 4 spliced pairs in the smaller rocket, but we couldn't quite get the last bottle to line up straight with the rest of them. We suspect that the bottle is bent ever so slightly when it was originally glued with the Sikaflex glue. Not wanting to launch a bent rocket, we decided to leave the rocket with just 3 spliced pairs.

To seal the end of the rocket we just use a normal bottle cap, but have a plastic reinforcement ring over the thread section to prevent the cap from flying off. We have tested this to 300psi (20.7bar).

All the tests were carried out with our high pressure (300psi) control panel as it was going to be the one we launch the rockets with as well.

Launcher modifications

Our medium launcher has nice long guide rails fitted, but we don't have a brass release head for it with stainless steel ball bearings that hold on to the nozzle. For the higher pressures we did not want to rely on the plastic tabs like in regular Gardena mechanisms. The nozzle was going to be aluminium to hold up to the pressures as well. Because our cluster launcher already has a built in brass release head with the ball bearings, we simply modified the base to accept the guide rails from our medium launcher.

Launch Day Report

• We arrived at the usual 8am timeslot and set up the launcher. The wind was already starting to pick up and heading in the wrong direction back over the rocket eating trees.
• We filled the Axion G-1 rocket with foam and 1.6L of water. We are using the jet foaming insert in the tornado coupling to produce foam on the way up. The new control panel allows us to fill the rocket quite rapidly to the higher pressures. We can get the rocket up to pressure in around 1 minute. We have switched to using a longer hose as well as a longer launch string. You just don't want to be anywhere near the rocket at those sorts of pressures.
• As we launched the rocket at 220psi, it very quickly became apparent that the chute was out a bit early ... well actually ...as soon as the rocket cleared the guide rails. It promptly decided it did not want to participate in the rest of the flight and fluttered down next to the pad. The rocket barely blinked as the parachute cord broke and continued to climb to 605' (184m). That wasn't too bad considering the parachute slowed it down somewhat.
  
  The nosecone was destroyed, but the flight computer survived, with only a bent connector and a broken power switch. Both of which can be easily fixed. The servo motor was destroyed and the 9V battery completely disintegrated. The camera and altimeter were located between the 1st and 2nd bottles so they were protected from most of the shock and survived, though the camera did not record the launch or it did not close the video file properly. The MD80 camera has an alloy body so I suspect it helped protect it.
  
  The top bottle was completely crushed, and even though the neck that was connected to the coupling was also broken, the coupling sustained no damage.
  
  We replaced the broken spliced-pair with the bent one and fitted a spare deployment mechanism to top. We replaced the broken parachute cord with a much heavier one and the rocket was ready for the next flight. I was surprised to see that none of the shroud lines were broken or pulled out of the canopy. They are simply attached through a hole in a piece of electrical tape stuck on the ripstop nylon.
• We set up the rocket again hoping that it was only a one-off. ... Nope... the second launch at 230psi was almost identical. The parachute fell out as soon as the rocket launched, but the heavier cord stopped the rocket from flying too high and the rocket landed close by without damage. We did get onboard video, but only partial altimeter data as the parachute cord slipped down the rocket and hit the altimeter power switch and turned it off.
• At this point we were quite stumped as to what was happening. We had used these two nosecones on quite a few rockets but never had this problem. At first I thought the acceleration must be too high forcing the parachute out, but that really didn't sit too well. The next theory was that the guide rail extensions were rubbing against the rubber band between the door and the servo and pulling it off as the rocket launched. This seemed reasonable and so we removed the top guide extensions.
  
  We also relocated the launch pad next to the rocket eating trees to give the rocket a clear path down wind. The wind was quite strong at this point and it was a 50-50 decision to launch or not. Not much more could have gone wrong at that point and so we decided to try again.
• As we started to pressurise the rocket the parachute fell out again. We stopped pressurising and I repacked the parachute. Started pressurising and the parachute fell out again. Dang ... just one of those days! So I repacked it, but did not arm the rocket. We pressurised it to 230psi again and only after the air stopped that I armed the rocket and then we launched it.
  
  The rocket flew great and went to 696' (212m). It was definitely noticeable how much more power the rocket had. I thought it was quite a reasonable flight considering it was bent and only 6L capacity.
  
  Even with a longer 8 second deployment delay, it drifted around 300m down range.
• Feeling pretty confident that we had solved the problem we set the rocket up again and this time pressurised it to 245psi (16.9 bar). We waited until the rocket was fully pressurised before arming the flight computer again. This is our highest launch pressure to date.
  
  The rocket went up beautifully to 744' (226m). Which we thought was a great result for this rocket considering it used a small 9.5mm nozzle and no launch tube. Running several simulations later that night we were surprised at the altitudes we were getting when they were predicting around 610'. I believe the foam has a lot to do with it as we have seen on a number of other occasions.
  
  From the onboard video it was noticeable that the rocket was spiralling quite a bit on the way up probably the result of the bent bottle.
• So what actually happened? It looks like the culprit is the modified launcher base. It has a built in non-return valve fairly close to the rocket. With the higher air pressures and flow rates it started hammering (vibrating) pretty badly and the vibrations travelled up the rigid rocket and caused the G-switch on the flight computer to detect launch and deploy the parachute.
  
  The reason we did not notice it on the first two launches was because the guide rail was holding the parachute door closed even though the servo had activated. Having removed the guides we were then able to see the fact the parachute was being deployed during pressurisation. You can hear and see the vibration in the on-board video as well.
  
  We did have an audible warning though which we didn't notice in the excitement of launching a new rocket. The FC beeps when it is armed. When it detects launch it stops beeping. We did not notice it had stopped beeping because we were further away than normal. I may update the code to make sure the FC is making a different tone if it is in the launch detected phase.
• Seeing how far the rocket was drifting we decided not to launch the bigger rocket (12.4L Polaron G-1) that was expected to go higher with a bigger parachute.
   

What's Next

• Next we would like to get the bigger rocket up in the air and see how it does. We want to launch it with a 15mm nozzle and launch tube. This in itself brings new challenges in designing for high acceleration.
• We are also going to see what the actual burst pressures are of the 90mm and 110mm spliced-pairs. This will give us a better idea what the safe launch pressures are.
• These flights showed that there is quite a bit of performance potential in these rockets. So we are going to pursue this further in upcoming, single stage, boosted and two stage flights this year.
• We need to buy more fiberglass as we've used up our 3m. I think we'll get a 10m roll next which is good for about 25 spliced pairs.
• We are also going to have a go at sealing the tornado tubes with either o-rings or soft seals so they don't have to be tightened so much.

Flight Details