February 15, 2004 notes
The custom honeycomb bulkhead for the next vehicle arrived,
and we have started prepping it. The
current cabin has a 40 diameter bulkhead midway up the cone with a fairly
sparse 2 thick aluminum honeycomb core.
We have pressurized the cabin to 18 psi without any problems. The new bulkhead will be going on the bottom
of the next vehicle, and is a full 63 diameter, so it needs to be stronger to
compensate for the larger span. We had
Teklam build the strongest panel they could conveniently put together, which
turned out to be a 3 aluminum core with increased density, and double ply
fiberglass facing sheets. This gives
significantly more margin than the current bulkhead, and only weighs 39 pounds.
The rolled aluminum cabin cylinder sections is slightly oversize,
so we need to build up the bulkhead perimeter some before starting to bond it
in. I got a big roll of 3 wide by 1/8
thick fiberglass, and we have epoxied two layers around the disc after filling in
the exposed aluminum core with flox. We
will test-fit on Tuesday to see if we need to add more.
I also got another large sheet of Nomex cored / fiberglass
faced honeycomb for our interior work. The
Nomex core avoids any chance of getting bits of aluminum lost in the cabin, and
the fiberglass faces come with peel ply, which lets us skip sanding work when
bonding to them.
We intend to have most of the necessary large scale
components for a new vehicle ready for the inevitable time when we crash the
Our custom motor drive boards and master cutoff computer
boards arrived. We had these contract
built, and had ten of them made, so we have plenty sitting on the shelf for new
vehicles and redundant computer systems.
The motor drive boards have 12 very high current bi-directional drive
circuits with opto isolation, status LEDs, and fuses. It is driven by a 50 pin IDC connector from
our AccessIO PC104 boards. Our existing
driver board only has four motor drives, which prevent us from using a fifth
lifting engine or going to a dual-quad engine arrangement. The current board also has eight solid state
relays, but we will just use motor drives on the new board for uni-polar
devices like solenoids.
We are potting all the stand-up components right now, so
after it cures I will replace the current board.
The current master cutoff computer is something that I built
up out of our old spare parts, and is ludicrous overkill: it boots Linux on a
PC104 stack to basically just listen to a serial port and hit a motor drive
when it stops getting sensible data from the main computer. The new cutoff computer is a simple little
microcontroller with four motor drives so it can sequence cutoff valves and
various emergency recovery options if the main computer fails.
We modified engine 0 some more by rebuilding the cold pack
with an additional 600 cpsi monolith and using a smaller area spreading plate,
but it still isnt working flawlessly.
It makes decent thrust on the test stand and doesnt quench, but it isnt
heating evenly, and it has a substream of clouds in the exhaust. We might be able to fly like that, but we
really want to get completely clear exhausts for visibility and best
We fired the 12 engine at some very low pressures, which
might be a flight option for us. All of
the catalyst necessary to make nine 12 engines should be arriving in under three
weeks, which is faster than ordering any new foil monoliths for 7
diameter. It warmed quickly and ran at
a very even pressure, but at only 40 psi of chamber pressure it didnt clear
itself out completely. This shows a
weakness of all our previous testing: We always opened the valves completely
when starting a test run, and many engines were briefly cloudy before
clearing. Trying to creep up the
throttle from idle just doesnt seem to work on the current engines. Slamming the throttle open isnt an option
for flight vehicles, unless we implement a hold-down system.
We have a couple more tweaks to try on the current 5.5/7 stepped
chamber engines, but we might be stuck waiting for new 7 monoliths if we find
that the stepped chamber is just problematic.
I have ordered a few thermocouple amplifiers for our data acquisition
system, which may let us get a little more insight into the operating
characteristics of the engines. Lots of
pressure taps let us finally understand all the thrust questions, so hopefully
lots of temperature taps will let us conclusively resolve the warming / start
The only catalyst that we have ample
amounts of fresh material of is the ceramic bead catalyst, so we did several
experiments with it.
Our previous experience with the beads was using them as a
hot pack on January 3:
cut open the ring engine and replaced the 600 grams of rings with 520 grams of
beads, which occupied a much shallower area.
This turned out to be much more restrictive than the rings, but showed
no change in behavior over the length of the run.
lbf from 265 psi feed, 208 psi mid, 83 psi chamber
removed 200 grams of beads (now 320), and the thrust went up:
lbf from 281 psi feed, 210 psi mid, 65 psi chamber (the sensor tube must have
been loose, this number cant be right)
We were afraid to pull any more of the beads
out, because the layer wasnt very thick, and any bowing of the supporting
plate would let them pull away from the edges and allow some gas to bypass the
This week, we initially tried making a cold pack with 375
grams of beads over a 600 gram ring hot pack.
We couldnt get this to preheat.
We tried to overkill it by adding a large section above the
existing section and filling it with beads, bringing the total to 1197 grams of
beads in the cold pack. This engine was
very interesting. It started making
very loud pops and bangs and intermittently stopping completely and
occasionally belching flame. Most
interestingly, the chambers were glowing a perfectly even red hot, all the way
up through the cold pack. When we
took it apart, the engine was a huge mess.
A lot of the beads had been pulverized, many internal supporting screens
and plates had been burned away, and bead fragments were cemented all through
the rings in the hot section.
The prospect of being able to sustain combustion in a single
pack section is very enticing, so we built another test engine with a single
block of 766 grams of beads. We left a
sizeable gap between the spreading plate and the bead catalyst, and mounted a
glow plug there. The hope was that
pulses of propellant would rain into the catalyst, and the vapors would ignite
on the glow plug and force combustion down through the pack to warm it throughout. We used the 1.25 nozzles so that the pack
would have a much lower pressure drop than with the big nozzle. It wouldnt be a bad trade to get a single
catalyst pack even if we had to go to increase our throat to chamber diameter
ratio from 3:1 to 4:1, but we couldnt get this to warm up. We might try again some time with propane
Even though this wouldnt burn reliably (it made a few
pops), it did seem to make a foamy mess about like the monolith cold packs, so
we decided to try putting a 320 gram bead hot pack section below it, with the
glow plug in the gap as is our usual practice.
We expected this to just work, because it was the same hot pack layout we
had successfully tested before, and the cold pack was probably overkill for the
amount of flow we were putting through it.
We were able to get it fairly hot on the preheat, but it quenched when
it was opened up. When we took the
engine apart, we had a very interesting sight:
the perf plate above the hot pack was very evenly burned from the
outside in. If we had run it much more,
we would have just had fragments inside, but we caught it at an ideal time to
see it nearly chewed through at the edges, grading to untouched in the center. The evidence is now pretty overwhelming that
the dust from the ceramic catalysts will catalyze the burning of solid stainless
steel as well as various gasses, which makes it rather difficult to make an
engine using it
The bead catalyst is
probably a dead end.
One other random thing we did this weekend was experiment
with using a dual stage venturi pump to draw a vacuum in the big tank as an
alternate loading method to pressurizing the drums. The venturi pump is pretty impressive, pumping air out many times
faster than our electric vacuum pump, but you do trade having to have a
compressed air or nitrogen source for it.
We arent sure which we like better yet. In any case, the new 50 extra-flex loading hose that arrived is
a huge improvement over the very kink-prone smooth bore hose we were previously