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.

We've been wanting to do this Zero-G (freefall) experiment for about 6 months now and finally managed to fly it this weekend. There have also been a number of other developments in the last two weeks so this is a bit of a long update. We'll cover the zero-G experiment first and then discuss, altimeter flights with and without foam, and wrap up with a design for a simple pressure switch.

Zero-G experiment

Inspiration

There have been a lot of discussions on the water rocket forums about gravity based apogee-detecting deployment systems. The theory goes that a weight hanging down will continue to hang down as the rocket flips upside down at apogee and the hanging weight will activate the parachute deployment system.

This works on the ground but does not work reliably in flight. This is due to the fact that the rocket and the hanging weight are essentially both in freefall from the time the rocket stops producing thrust. This freefall extends from burnout, through apogee and all the way back to the ground. The only other forces acting on the rocket are caused by drag through the air, which decreases as the rocket slows down near apogee, and increases again as the rocket accelerates.

We wanted to demonstrate what happens on board a rocket under these freefall conditions. If you recall Mike Melvill's famous sub-orbital flight on SpaceShipOne and in particular him releasing a hand full of M&Ms, watching them float around the cabin in Zero-G. The video can be seen here the M&Ms get released about 1:30 into video. AGR performed a similar experiment a few years ago.

So we decided to replicate the same experiment on one of our water rockets. Compared to Mike's 3 minutes of Zero-G, we ended up with only a few seconds, but long enough to see the same effect.

Experiment Setup

One of the keys to making this experiment work was to allow the rocket to freefall as long as possible. This meant opening the parachute well past apogee. We used the flight computer's ability to set a longer time delay from launch to parachute deployment.

We used water only (not foam) on these flights as we wanted the rocket to stop producing thrust as soon as possible after takeoff in order to enter freefall earlier.

We used our Hyperon Standard Platform rocket in a three bottle configuration as the rocket motor with a long experiment payload attached on top and finally the recovery system nosecone on top of that.

The experimental payload was constructed by splicing three 1.25L bottles together with the bottle bases at either end. Since the experimental payload is not pressurised the joins only needed minimal overlap.

A small video camera was attached inside the top of the payload looking down towards the other end of the payload. We also emptied a packet of M&Ms into the bottom of the payload. (See diagram on left)

Preflight Experiments

Before flying, we first performed a number of ground experiments to see if the concept would work. We tested the focal length of the camera to see if the M&Ms would stay in focus or whether we would need to add a lens to make them sharper. In the end we decided that the amount of blur was acceptable and decided to go without a lens.

We also subjected the M&Ms to forces on the kitchen scale to see at what acceleration they would crack. They held up well to over 100Gs so that wasn't going to be an issue.

We also tested the effect simply by throwing the payload section with the camera and M&Ms into the air and catching it again. This showed the Zero-G effect although only briefly.

Flight Day Events

• The Graviton rocket was the first off the launch pad for the day. We set the parachute deploy to 2 seconds later than normal. The flight went well and the rocket pitched over well at apogee. The parachute blossomed about 10 meters above the ground, but because the parachute string was too far back on the rocket, the rocket continued to point towards the ground and ended up hitting pretty hard. The nosecone suffered minor damage, but could not be flown again on the day.
  
  The payload video was interesting although not unexpected. The M&Ms flew towards the camera after burnout and remained there (actually behind the camera) for the rest of the flight.
• We repaired the rocket and added a bulkhead around the camera so the M&Ms could not fly past the camera. This had the effect that some of the M&Ms covered the lens during the flight.
• On the next day we flew the Graviton rocket again three times to see if we could get a better video of the M&Ms. The videos always show the zero-G condition just after burnout, with the horizon in the background pitching over much later as the rocket goes through apogee.
• One of the flights was unsuccessful because the camera was in the wrong mode to record video.
• The rocket behaved well on all three flights and with minor damage that will be repairable quickly. The one thing we were also happy about was that although the rockets had a few harder landings, the now glued servo motors remained intact.

(If the video does not play, try the latest Flash player from Macromedia)

Experiment Conclusions

• The experiment clearly demonstrates that after burnout all unattached things inside the rocket fly forward and remain there until impact with the ground. This is due to the induced drag force caused by the rocket flying through the air. This has the implication that most gravity based parachute deployment systems will actually activate at burnout rather than at apogee.
• We were hoping to see more of the zero-G effect of the M&Ms floating around for longer, but we did not know really what would happen until the experiment was flown.
• We now have a number of ideas on how to improve this experiment, and get the M&Ms to float around longer but more on that next time.
• M&Ms on a hot day inside the rocket can become quite soft. After the last flight a lot of them were pulverised.
• We want to fly this rocket on a less windy day where the turbulence on the rocket will be reduced. We also want to reduce the rocket spin as much as possible.

Altimeter Flights

After more than a year of flying water rockets we finally bought a couple of altimeters so that we can get some direct feedback on rocket performance as we make design changes. The altimeter is the Zlog Mod 4. by Hexpert Systems We are quite impressed with the accuracy and functionality it has. We fitted the altimeter to the newly rebuilt J4 IIIb rocket.

One of the main reasons for choosing this altimeter was that it can be started on the ground and does not need a certain gain in altitude to start recording like some others. With some of our slow take-offs this wouldn't have triggered the other altimeters. At 8 grams and powered by the flight computer battery, it has minimal impact on rocket performance.

We mounted the altimeter in the nosecone section and padded it with foam. We used the altimeter's servo pass through feature which meant hooking it up was trivial. We just plugged the altimeter to where the servo plugs into the flight computer and we plugged the servo into the altimeter. There is a small hole near the altimeter in the payload section to allow air pressure in and a small door covers the altimeter so that we have access to the altimeter for downloading data.

Flight #1

We only flew the rocket twice because of the strong wind. The rockets were drifting a long way down range and towards trees and the river.

We flew the first flight with water only and at 110psi. The parachute opened a little past apogee. This can be seen in both the video and altimeter data. The altimeter data is shown below:

(click to enlarge)

The rocket reached 321 feet ( 98 meters )

Because the initial launch elevation and final landing elevation are more or less the same in real life and the altimeter zero point was set a little above these, then to get a more accurate altitude we use the midpoint between the before launch and after landing altitude measurements and added this to the altitude.

We then entered the launcher and rocket parameters into Cliff's simulator and it predicted an altitude of 327 feet. Which is in very close agreement. This gives us further confidence in both the altimeter and simulator accuracy.

Flight #2

Our second flight had an almost identical configuration except we added a little foam 40mL, (the water amount was the same), pressure was increased from 110psi to 120psi and the water/foam mixture was set up to use the Jet Foaming technique.

The altimeter data is below:

(click to enlarge)

The rocket reached 483 feet. ( 147 meters )

The one thing that really surprised us was that the simulator predicted that the rocket should have reached 369 feet ( 112 meters ) with the added pressure. This is a very significant difference and is continuing to support our suspicions that foam actually generates a significant performance boost over straight water. The flight also looked very high in the video. From the altimeter data we also get a flight time of 31.8 seconds.

We have already started seeing evidence from previous flights of the foam's performance. However, these two flights aren't enough to provide reliable evidence and so over the next few weeks we are going to try to fly a lot more comparison flights to see what the real difference is between water-only and foam.

Flight Details

30th September 2007

1st October 2007

Pressure Switch

We designed and tested a simple pressure switch in the last two weeks as well. The purpose of such a switch is to detect when the rocket stops producing thrust. This is useful for initiating staging or combined with a delay to deploy a parachute. The whole thing can be built for around 10c. This switch carries out the same functionality as Trevor's TDD without the need to make a large hole in the pressure chamber that can weaken it. The pressure switch can be used at either end of the bottle.

A pressure switch has the advantage over pure launch detect switch because you can vary the amount of water, nozzle size and pressure in the rocket without the need to change the delay for each combination.

The switch activates at around 10-20psi. Actually the switch resistance changed as pressure increased, although not linearly enough to be used to measure pressure. But works well as pressure/no pressure switch. (Fully activated is around 200 ohms, without pressure it is several mega-ohms)

The whole thing is made out of a button cut out of the rubber membrane from an old remote control. The button contact is made from a section of the remote control's PCB. The button and PCB are just tightly sandwiched between two pieces of aluminium with a couple of screws. The sandwich is sealed with epoxy except where the rubber button pokes out. The rubber flange around the button provides the seal.

Atmospheric pressure is just trapped underneath the button so that the entire unit can exist inside the pressure chamber except for the wires sticking out. Since this was only a prototype I just epoxied the wires into a lid. Eventually the whole unit will be glued to the lid and not dangling by the wires as shown here.

The whole thing weighs 10 grams but you could probably shave off half that without too much fuss.