December 31 2002 and January 4, 2003 Meeting Notes
We finished assembly of the nosecone (1/8 roiled aluminum)
and engine bulkhead today. The next
time we do this, we are going to invest in some proper leveling tables and laser
sights, because getting 2 diameter tubes straight by eye is pretty hard to do
repeatedly. If this vehicle tests out
exactly as it is supposed to, we will be moving to full X-Prize size vehicles
next. If, as is likely, we wind up
planting this one during the test series, we have a spare cone, tank, and
enough tube to build another one just like this in fairly short order.
The CNC mill cut the ½ thick aluminum bulkhead, the 4
degree engine canting shims, and a custom tooling block that aligns and locates
the drill for creating all the side-bolt holes in the tube and bulkhead. It is turning out to be a very good purchase
We were prepared to make a large fin assembly out of
honeycomb core composite boards, but we have decided to fly finless, relying on
the control system for stability during the burn. This saves a fairly notable amount of weight, forward drag area,
and fabrication work. The rocket will
be immediately unstable at burnout, so we will rear-eject a drogue chute large
enough to stabilize the vehicle either upon sensing an attitude change, or
based on a timer. A Kevlar drogue
should be able to be ejected while the engines are still firing without harm,
given the relatively cool temperature of the monoprop exhausts. The drogue will be held by an electronic
release, which will allow it to pull the main chute out after it has arced
over. We are going to investigate using
electronic trunk releases as a cheap-but-sturdy possibility. Popping a stabilization chute at burnout
will severely limit the altitudes the vehicle will attain with this vehicle,
but that doesnt bother us at all, and should actually save us a lot of
recovery work. When we make a
space-shot vehicle, the trajectory will be tailored to burn out at a very high
altitude, mostly negating the issue.
This vehicle should still be capable of supersonic flight, given a full
load of peroxide.
We are considering making the very first flight with the
drogue pulled out by a static line shortly after liftoff, which would limit the
altitude even more (it would be towing the drogue through most of the burn),
but would let us factor out some uncertainties with the flight, and still get
data on some other things we care about.
This assembly was 200 pounds almost exactly, but that
doesnt include the master cutoff valve/manifold, electronics, parachute, and
nose cap. Still, even with everything
on, it will be lighter than the previous vehicle with the single engine and
We have an Ashtech G12-HDMA speed/altitude unrestricted GPS
with 10hz update rate, but it was quite expensive (> $7,000), so we arent
going to put it in the vehicle until it has been successfully flight
tested. This unit is used in various
space and missile systems, and is export controlled.
Still to do:
I need to get our new Ampro PC104 system configured, because
the WinSystems SBC was killed in the last crash. We could use the Mach-Z PC104 system that we still have, but the
Ampro has soldered-on RAM, which removes one possible vibration failure point
for us. I was unable to find anything
faster than a 486-133 with soldered memory.
The long SIMMS on modern systems seem like an invitation for failure.
New redundant power supply board, through-hole-plated driver
board, and master cutoff watchdog board
Engine / release cabling.
It is tempting to run these actually through the propellant tank, but we
will probably epoxy a little conduit on the side.
Fill connector and pressure gauge on bottom bulkhead
Crushable nose cap fabrication and testing.
Build and test three more 5.5 engine catalyst packs
Temporary landing pads for hover testing
Drogue cannon arrangement
Main chute release mechanism
Drop G Forces
We tested a 50G accelerometer for landing impact data
acquisition, and learned a few things.
Somewhat surprisingly, you can easily give the sensor 40+ Gs by just
making a fast pitching motion with your hand.
The sensor I got only has 100hz bandwidth, which isnt enough to capture
sharp decelerations. If you just bang
the sensor down on a flat table, it doesnt register much of a signal at all,
because it happens in too short of a time span. I am probably going to have to get a vibration / shock
accelerometer, which can give much higher bandwidth, at the expense of not
measuring constant accelerations.
We upgraded to one of the professional data acquisition
systems from Dataq a while ago, but I am still using WinDaq-Lite for data
collection, which limits us to 240 hz sampling. If I get a mult-thousand Hz bandwidth sensor, I am probably going
to write my own acquisition software, because paying $2,000 to Dataq to
basically just change a constant in their software seems like a ripoff after I
paid >$2,000 for the hardware. Im
not thrilled with the screw terminal breakout block on the system, either.
We strapped Russ and the accelerometer into the
custom-foamed couch we made last week, and let him report how a few impact G
tests felt. The foam has almost no give
to it, so you stop quite abruptly when it lands flat. We dropped him a few times, and there is a big difference if it
landed on one end and pivoted down, or landed completely flat. The last completely-flat drop registered 15
G, but was probably higher due to the sensor bandwidth issue. Russ said it was definitely a pretty hard
impact, but non-damaging. 50Gs is a
test point for a lot of race car safety equipment that is supposed to function in
We have been doing our decelerator estimates based on 10G
deceleration for under a half second, which seems reasonable. The current vehicle is only going to have
about 6 of crushable nose cap, but the big vehicle is going to have over 8 of
cone to crumple up ahead of the cabin, so it should tolerate even fairly high
parachute decent rates.