July 14 and July 17, 2001 Meeting notes
In attendance:
John Carmack
Phil Eaton
Russ Blink
Darin Smith (17th)
VTVL Flight Test
On Saturday, we intended to do translation maneuvers with
the VTVL.
A couple times during the week as I was working on it, I
experienced a computer lockup. I had been attributing them to low battery
power, so I made sure it was fully charged before the flight test, but I was
worried enough about it that we decided to just start with another tethered
hover test instead of the planned translation maneuvers.
The computer was locked up when we were about to do the
water test. At that point, we really should have canceled the flight
tests until we figured out what was wrong, but we decided to press on.
I plugged an external power source into it, and got the
computer back up, then it seemed to be running fine on it's own battery.
It ran the water test fine, but when we loaded it with peroxide, 1/3 of a
second after liftoff (as determined by the telemetry logs) the computer died
again, leaving an attitude engine on. It very rapidly did almost a
complete flip.
We handle this situation a lot better now than when it used
to happen on older versions, because we have dedicated connectors that allow us
to easily vent the system without the computer, but it still ended the testing
for the day.
One of our foam support plates was bent out of shape, but
nothing structural seems to be hurt.
One of the FOG supports came loose in the electronics box,
and one wire broke off at the terminal, but nothing was damaged.
That night, I finally tracked down what had been causing the
problems. I had added two more users of +5V regulated power over the
week, and that had forced me to double up ring terminals on terminal
posts. The terminal with the CPU power and one other user was loose, so
rattling it around could cause the CPU to momentarily lose power. Doubled
up ring terminals can sort of wedge together and feel tight, even when they
aren't.
I learned my lesson -- I put on a longer terminal strip with
proper jumpers, and am now limiting myself to one ring per terminal at all
times.
Peroxide Drama
When I got into work today, there was a message from the
warehouse manager at Rinchem saying that our drum in peroxide storage had fluid
pooled on top, and had leaked some over the sides.
I was quite alarmed, but they said that the drum wasn’t
warm, so it didn’t seem to be a critical danger. I called Michael Carden at X-L Space Systems (our peroxide
supplier), and he contacted Rinchem with some additional handling instructions.
At this point, my theory was that because we had left the
polyethylene hand pump sealed in the drum, perhaps slow pressure buildup could
have pushed peroxide out the pump instead of venting through the relief
valve. This would have been somewhat
disturbing, because in our original discussions with FMC, they recommended
never removing a pump once you put it in a drum.
Michael mentioned another possible cause – some of these
stainless drums had had problems where a small amount of peroxide had gotten
between the PE liner and the drum wall, which caused the liner to inflate and
push peroxide out of the drum.
Phil and I drove over to Rinchem to investigate. When we got there, the drum was still cool
to the touch, but there was peroxide covering the top of the drum, all the way
up to the edge, and a little bit of runoff over the side into the containment
palette.
We suited up completely and started cleaning it up. We didn’t have a great way of getting the
peroxide off the drum. We ladled as
much off as we could with a small cup (dumping it into a bucket of water), then
we started sopping it up with a rag, squeezing it out underwater in a big mop
bucket.
We tried pumping some peroxide out, but the pump wasn’t
working very well, and it looked bent internally.
Unscrewing the pump was difficult, but when we got it loose
we could tell that the pump was pushed hard towards the center of the
drum. When we removed it, we could see
the PE liner ballooned inward just like Michael mentioned. We took the corrective action he recommended
– prick it with a pin and let the oxygen vent out.
Everything seems fine now, but I need to buy another hand
pump.
Electronics Box
I got the DC motor drive working properly, so we have all
the connections for the big ball valve throttles now.
Russ sanded off 1/8” from the bottom of the box foot
moldings so we don’t have to struggle so much to get it into the VTVL next
flight.
I have some flanges coming in the mail so we can get the box
rigidly secured to the frame with thumb screws next time. We had been forced to the soft foam mounting
because of the vibration sensitivity of the Gyration gyros, but now that we
have the fiber optic gyros, we can bolt the box directly to the frame.
The convenient little adhesive back wire stays don’t seem to
stick well enough in the box, so I am moving to bolted loops for everything.
Engine Tests
media.armadilloaerospace.com/2001_07_17/jul17_01.xls
We tested the ½” throat (manned vehicle attitude) engine again
today, with two changes:
The perforated metal injector plate is McMaster 92315T121 instead of 92315T101 , which is only 9% open instead of 21% open, and is a slightly heavier
gauge. We believe that we need more injector
pressure drop to smooth out the performance.
Instead of a drilled piece of brass as the catalyst
retainer, we trimmed a piece of heavy gauge perforated metal, which offers both
more open area, and smaller holes. We
had experienced some foam cell breakage with the larger drilled holes, and the
small total area was uncomfortably close to the nozzle area.
The first test was 500 ml of 80/20 mix at 500 psi. It started out extremely smooth and level,
probably better than anything we have ever had, but towards the end it
roughened up a bit.
I added a 200 msec warm up pulse back to the test script for
future runs. Even though the catalyst
packs don’t require preheating, it is still good to burp the plumbing.
We then tried a long duration run, with 2000 ml of 80/20 at
an initial 600 psi. It was moderately
rough, but the roughness diminished as the pressure blew down, which is similar
to what we saw with our low pressure tests last time, but sort of the opposite
you expect of normal rocket injector behavior.
We then tried another 500 ml. 500 psi run exactly like the
first one. It was moderately rough the
entire run.
We took the motor apart to see what had happened to the
various parts.
The perforated steel retaining plate worked perfectly – no foam
had been pushed through, and it wasn’t deflected in the slightest.
The injector plate behaved just like the last (more open)
one, with good sealing and a slight residual dish to it.
The catalyst pack had been compressed about 1/8” away from
the injector plate, which probably accounts for the increased roughness after
the first run, because peroxide wasn’t being forced immediately into the foam
after passing the injector plate. I
think a reasonable argument could be made that it is harmful to have more
catalyst discs than you need, because a larger stack will be easier to compress. The perforated metal injector plate does a
marvelous job of spreading the peroxide evenly over the surface of the pack,
and we know that 15 1” diameter discs were capable of catalyzing at least 50
pounds of thrust, so these 1.5” diameter motors should be able to get by with
only six or eight discs.
The unexpected thing that we found was that the top foam
disc was COMPLETELY stripped of silver.
Russ thought it was an un-plated nickel disc. Our peroxide concentration was less than 85%, so it shouldn’t
have been melting it off, but it might have been a mechanical stripping
action. No silver was deposited in the
nozzle, so it must have just been deposited on the rest of the pack.
The next few discs also had some stripping, but the second
half of the pack looked completely untouched.
I probably should continue to try to buy pure silver foam
from Porvair, but they haven’t been very responsive. Phil found another company that would make silver foam, but they
want a $5000 setup fee. Pure silver
foam would completely eliminate stripping as an issue.
The basic engine design we have now seems to be a reliable
performer, but I would like to see if we can get the pack to perform continuously
at the smoothness level of our initial test.
Our rules of thumb for engines are now:
nozzle exit diameter of two times the throat diameter (4x
expansion ratio)
15 degree exit cone
45 degree converging cone
chamber / catalyst pack diameter of three times the throat
diameter
9% open 35 gauge perforated metal injector plate clamped
between the engine flange
50%+ open 18 gauge perforated catalyst retaining plate that
sits on a lip in the engine
8 to 16 discs of foam catalyst left uncompressed between the
injector and the retainer
We are going to try to build a 1.5” diameter throat engine
to this spec next, which should be over 600 pounds of thrust. We may need to add a heavier gauge
reinforcement behind the thin injector plate, but we will try it without any
modifications first.
Russ has finished the second ½” throat engine, so he has two
more to go before we can hop the big frame on just the side engines.
Once the big engine is done, we will be strapping some form
of makeshift crash test dummy to the big frame and flying it under the same
conditions as a manned flight.
News
