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.

Photo: Andrew Burns

Photo: Andrew Burns

Date:  29th January 2011
Location: Doonside, NSW, Australia
Conditions: 27 Degrees C, light winds mostly cloudy.
Team Members at Event: PK and GK

We were happy to finally get the G2 rocket up into the air today. It's been a long process to come to this point, however, the flight was not as successful as we would have liked. Though disappointing we did learn quite a lot about launching a rocket of this size. We also have some good on-board data from the flight, which will help us with the next series of G2 flights.

We had finished the rocket earlier in the week and were ready to launch when we received an email from Alex from (http://ckona.rocketworkshop.net/) in Ukraine with a good suggestion to fill the rocket with water using an external water tank that gets pressurised to push the water in. (http://balancer.ru/_bors/igo?o=balancer_board_post__2348977). This is a concept that has been used by others before, but I hadn't given filling the rocket with water much thought and expected us to fill it on it's side as usual. We normally tilt the launcher over to load the rocket and then stand it up. But with this size rocket that was going to be very difficult to do. Alex had a very valid point.

We didn't have enough time before launch to set up the pressurised water tank, so we opted for assembling the rocket on the pad. We put half the rocket on the pad and filled it with water. We then prepped the parachute into the nosecone and screwed the other half of the rocket onto the bottom half. We had to then assemble the upper part of the launch tower around it. It ended up being relatively simple, but in the future we should use the pressurised water tank.

We also modified the Acceleron V cradle to make it easier to transport this rocket in sections.

Flight Day Report

The whole launcher and rocket took about 1.5 hours to set up.

After the rocket was set up we started pressurising it. As we reached about 220-230 psi I could hear a leak somewhere on the stack. With that much pressure you really don't want to get too close to investigate, so dad kept filling up until ~250psi, but when the tank was turned off, the pressure had gone down to about 220psi by the time we were about to launch. So we re-filled it, but the leak continued, and the final launch pressure would have been closer to 220-230psi?

We launched regardless as there was enough pressure in the rocket to operate correctly. If the pressure had dropped much below ~190psi we would have had to abort the flight as the parachute would have opened late and there was a danger of it ripping off.

The first part of the flight went very well as expected with a peak acceleration of 15.6G, but right at peak velocity ~88m/s (317 km/h) and 1.17s into the flight the nosecone came off and the parachute deployed 0.3 seconds later. Not surprisingly the parachute also let go of the rocket. Later when we looked at the nosecone and parachute line it was clear that the edge of the fiberglass nosecone had cut through the cord as the parachute snapped open. The nosecone had about a 1cm zipper in it. After the cord snapped the nosecone separated and fell back down on it's own and hence is mostly intact.

From the on board data we can see that the rocket lost about 4m/s when the nosecone came off and hung on by its string, and another 9m/s when the chute opened. This means the rocket lost about 47km/h in speed as a result of the event.

Acceleration and velocity data from Craig's flight computer. Click to enlarge

The rocket no longer had a nice elliptical nosecone but a flat blunt one which added a lot more drag. See the following for more details: (http://www.aerospaceweb.org/question/aerodynamics/q0151.shtml)

The rocket arced over at 668 feet (203m) and crashed heavily 14.1 seconds after launch. Due to the lower launch pressure, the parachute deploy anomaly and the higher drag, the rocket did not achieve its predicted altitude.

Z-log altimeter plot

What Happened?

We are still investigating why the nosecone deployed early. We suspect it was just enough uneven air pressure on the nose that made it slide off sideways. The lip that keeps the nosecone from sliding is fairly small and with the added pressure of the springs wanting to force the nosecone off that's perhaps all it took. There may have been a part of the parachute caught under the lip.

Another possible issue was that the timer could have activated too early, and even though I was confident the time was set correctly we wanted to check the EEPROM setting. The PIC was dislodged from it's IC socket and all the pins were bent on impact. Straightening the pins and inserting it in the PIC programmer we confirmed that the time was set correctly.

It is possible that the timer triggered at an earlier event while the rocket was still on the pad, and happened to expire about 1 second into the flight. Though early deploy is unlikely to be the cause as the nosecone came off at peak velocity shortly after burnout. This would be too much of a coincidence for the timer to fire at the same time.

The deployment mechanism was in the deployed position when the rocket was found, but that does not mean it activated early as it may have activated 9 seconds into the flight as designed.

Both of the o-rings holding onto the latch were found intact and still attached to the nosecone so o-ring failure is also no the explanation. It is also possible that at least one of the o-rings slipped off the latch.

Damage

The damage to the rocket was quite extensive and included:

• V1.6 flight computer (board snapped) - some components are salvageable
• Nosecone tip - fixable
• Bottle cap reinforcing ring - replace
• 3 x fairings and camera and altimeter housings - replace
• 9V battery - replace
• Servo motor - replace
• Deploy mechanism crank - replace
• Remote arming switch - fixable
• Fiberglass nosecone base - replace
• 3 x reinforced spliced quads - replace

However, the following survived with no damage:

• Tail spliced quad and fins
• 3 x tornado couplings
• Altimeter - (damaged battery)
• MD80 clone camera
• Craig's Flight computer
• 2 x 350mA Lipo batteries
• 16mm nozzle
• Parachute

The most critical components such as the altimeter, camera and Craig's flight computer survived which is what we were happy about.

Other Observations

When the rocket crashed the power switch next to Craig's flight computer was bumped and cut the power to the FC. This meant we lost some config data and first 20 samples. These 20 samples included a part of the launch tube boost phase so not really important.

When we were packing up the launcher I found the nozzle seal lying on the ground next to it. I unscrewed the nozzle from the rocket and sure enough the seal was not there. That was odd because if it had come out during pressurising we would have seen the water leak and it would have stayed threaded on the launch rail. When the nozzle is tightened part of the seal protrudes on the inside of the bottle. I suspect the shear pressure on the seal while all the water is rushing out is probably what pulled it out through the nozzle.

The rocket also looked like it had a very slight bend in it in the middle coupling, but it did not look like it affected the flight too much. This may have partially contributed to the uneven airflow.

Leak

We are currently tracking down the source of the leak we saw on the pad during pressurisation. There are a number of areas where it could have originated: Top cap, 3 x tornado couplings, bottle/nozzle seal, nozzle/launcher seal, launcher release head seal or hose.

We are pretty confident that both seals around the nozzle were good because we heard a hissing noise and no water. Two days before we did a whole rocket pressure test to 100psi to check for leaks and all was good. We only uncoupled the middle tornado coupling so when the rocket was re-assembled on the pad it could have potentially been the cause.

The other leading candidate is the hard rubber seal on the release head. We have not tested this one to 250psi and will need to check that but because it is fitted with a launch tube we need to put a long high pressure rocket on it first. We'll do this in the coming days.

We can't check if the cap leaked as it was destroyed.

All three tornado couplings are intact with the bottle necks snapped and still inside them. There are no obvious signs that the o-rings slipped out as has been seen before with softer o-rings (http://www.aircommandrockets.com/images/day89/Day89_05_s.jpg). We are using hard o-rings which should eliminate that.

There is one additional anomaly we saw in the altitude data from Craig's flight computer that could potentially be explained by pressurised air escaping into the fairing from the tornado coupling. The computer with its barometric sensor were inside the fairing with only a small hole for the switch access letting out any air. This fairing was right over the middle join that was assembled on the pad.

This tornado coupling and the release head seal are the two likely candidates.

What Worked

Despite the damage, we learned:

• The spliced quads work well. This was their first test flight.
• The rocket was stable in flight.
• The fins fit through the launch tower ring braces during launch. There is only a small clearance.
• The fin attachment method worked on this size rocket.
• The rocket didn't explode on the pad at the 250psi.
• Locating the electronics further down on the rocket prevented them from getting damaged.
• The rocket did not bend severely during the high G-launch.

What's Next

We are still deciding whether to go back to a side deploy mechanism, or try to get this nosecone working more reliably. There are advantages to both approaches. We are likely going to increase the strength of the parachute cord, and add a shock cord to reduce chances of ripped off parachutes. We still have 3 more new spliced quads we can use, 2 of which need to be sanded and pressure tested. We are also likely to use a different setup for switching power on and off to electronics to prevent power interrupts in case of another crashed rocket.

The nozzle is going to be re-made to use o-rings instead of the flat seal.

For the next set of flights we are going to add a second redundant recovery system. There will be a weight penalty initially, but when we get the main deploy working properly, then we will be able to remove the backup for better performance.

Acknowledgements

Thanks goes to Craig and Andrew for helping us set up and launch the rocket.

Day 100 - Highlights

Flight Details