Sunday, March 21, 2010

Blue is dead...

We had a perfect tether flight on Saturday, followed by a bad free flight.
The vehicle started well, but as airspeed built up it started to spin.
It got to about 1Kft then started down. It went unstable probably due to the high rate of spin.
and started to leave the area of the pad and was heading toward the spectator area so I aborted. Both the software driven and RC only vent abort worked.
The vehicle is totaled.

The micro sd flash chip from the vehicle physically looks ok, but gets very warm when power is applied so I've gotten no data from the vehicle.

I had a camera pointed at the bottom of the motor with a good view of the vanes and if I can recover data from the video camera SD card it should provide some information.
The camera was destroyed on impact, the SD card is fine but the camera did not close the video file before impact. So it shows zero length.

Any advice on how to fix this?
My approach:
I've got some utilities for manipulating bare SD sectors that I use for my data logging cards.
Using this I wrote a utility to copy all the sectors off the Cameras SD card into a big file.
So the original card is safe and I won't mess with that.
I then wrote a short utility to copy the exact same sectors contents back onto a equal sized SD card. I'll then run chkdsk on this duplicate SD card.

One rub is that while the SD cards are the same brand and size they report slightly different numbers of physical sectors, will this mess up the FAT table accounting? The card I'm copying to has more sectors than the one off the camera.

The cards is formatted as FAT16 so I can start tracing clusters by hand and reassembling file chains, but its been along time since I did FAT table stuff by hand with a hex editor.
Any recommendations on good tools to do this?

If that does not work,
any other ideas on recovering the video data from the SD card?

Tuesday, March 16, 2010

Getting ready part N..

I've made my cat pack. I've built new tethers, the only thing left on my list is re zero the one vane I had to take apart and do final electronics checkout. In the next day or so I'll modify the software to match what I'm flying in simulation.

I'll probably leave for FAR Friday night and do a tethered flight very early Saturday. I've done a lot of simulations and a flight to the 1200 ft limit of G airspace. (1000 ft to give me some margin) and back to the pad is easily within the most pessimistic performance parameters. So if the tether flight goes without problems I'll make that attempt on Saturday.

Higher than that and I'll need FAA authorization to enter controlled airspace. (Which they are unlikely to grant me until I actually stop being lazy and formally ask for it. )

When I get that I'll try the flight to 11K ft and back to beat DCX. With out some aerodynamic drag on the way down it has ~3% margin. As a result I've ordered a parachute from Fruity Chutes (blue of course) and I will setup to deploy the drogue at apogee and basically fall back to 1200 ft or so before powering up to land. The purpose of the chute is to keep the vehicle upright and limit the speed to 60mph or so. This will either require a very calm day or compensation for winds aloft. I need to mdoel this. I can just use forecast winds, or I have to study the FAA rules to learn what I need to do to release a small balloon to measure the winds right before launch.
(A small Xbee transmitter and GPS cost less than $100 and is very light so I can actually fly a small instrumented balloon.)

Rough calculation says 11K ft to 1K ft at 60mph takes 113 seconds. So a 10mph wind would cause 1600 ft of displacement. I'll need to be very close on the wind calcs to land back on the pad. I might be better off to fly from the lake bed near FAR.

Saturday, March 06, 2010

Working on Blue and....

Last week I fixed the landing gear on the blue ball in preparation for taking it to spaceup last week for show and tell. (You can see a picture of it there on its new gear here:

Last night I worked on building up a new cat pack. I started with the cat pack from silver and added some new disks. The top third of the cat pack was stripped and dead, I'm not sure what is going on. If I'm overheating the cat pack I'd expect the bottom third to be bad, not the top third. This leads me to believe I have some kind of contaminate in the h2o2. The new pack is ready to install.

My prep list looks like:
  • Remove the bottom half of the motor.
  • Grease lube and inspect bearings on the vanes.
  • Install cat pack in motor.
  • Replace seals in all the sanitary fittings. (This is preventive)
  • Reassemble the motor.
  • Replace the GPS antenna cable. (It has a kink at one of the ends.)
  • Inspect the wiring.
  • Build new Tethers.
  • Do a full electronics checkout.
  • Charge all the batteries . (computer, actuators, abort rx, telmetry box, laptop,2x rc transmitters, 3x cameras )
  • Do a tethered flight (on the 20th)
  • Do a free flight to about 1K ft. (maybe on the 20th)
  • Do a higher free flight???
I expect to finish all but the last four this weekend.

I have almost no traditional HPR experience with things like recovery high speed airframes etc... to remidy that I've been contemplating building a really simple H2O2 monoprop out of 4" or 6" aluminum tubing. Something that could go supersonic and reach to 20k ft or so. I'd use conventional dual stage recovery just like the big HPR guys do. This is something I can do with today's budget. Making the same basic vehicle bi-prop would make it capable of 100K ft.

It would probably be three increasing complex projects:
  • Unguided simple blow down mono prop to learn recovery.
  • Fin guided pressurized mono prop.
  • bi prop.
One of the really cool things I saw at spaceup was Ventions LLC small 100lb thermally
decomposed biprop h202 and RP-5 motors. Dr London had some cool video of it fireing on his phone. The motors were interesitng for both the construction and the thermal decomposition.
The were constructed out of stacked photo etched plates diffusion bonded together.

Getting a reliable thermal decomposed h2o2 motor to work would also be a really cool project.