January 11, 2004 notes
We cut apart the most recent welded ring catalyst engine and
added a third layer of 200 grams of rings and replaced the 2.2 nozzle with a
1.7 nozzle, giving the same 3:1 catalyst to throat ratio as the big motor. The big motor should be about 5.5x the
thrust of the small motor, going from a 1.7 throat to a 4 throat.
The motor configuration is: 204 x 0.040 hole spreading
plate, ten 20 mesh screens, 2 x 1 thick 600 cpsi monoliths, 3/16 support bar,
1/8 perforated flameholder, glow plug volume, 1/8 perforated plate as a top
ring holder, 3 x 200 grams of rings supported by ¼ thick perforated plate with
a 10 mesh 316 screen to keep the rings from pressing into the holes.
We made a very long burn to see if maximum heating would
cause the rings to collapse. It ran 65
seconds at 400 lbf thrust with no problems at all, although parts of the
chamber were glowing yellow hot. We did
two more runs for calibration at different pressures:
775 lbf at 425 psi tank, 300 psi mid, 248 psi chamber
410 lbf at 232 psi tank, 184 psi mid, 149 psi chamber
There is no sign of warping on the thicker perforated
We assembled the big motor with similar pieces:
In the top chamber: 1688 x 0.032 holes spreading plate, 10
20 mesh screens, a single 2 thick 900 cpsi monolith, a milled radial support
plate with a 10 mesh 316 screen between it and the monolith, and a 1/8
perforated flameholder welded to the support plate.
In the mid chamber: a 1/8 perforated plate as a top ring
holder, then three sections of approximately 1000 grams of rings supported by a
¼ thick perforated plate backed up by a cross of ½ thick stainless square
The mid chamber is flanged to bolt to the top chamber and
the nozzle. Assembled, the engine is a
huge pig, weighing 125 pounds. We want
to wind up at about half that on the final engines. All welded construction and more efficient supports will get us
most of the way there, but we may wind up going with a 6 tall by 12 ID by 1/8
thick inconel tube for the main chamber.
We made a couple short tests at our shop:
965 lbf at 125 psi tank, 89 psi mid, 66 psi chamber, this
run showed a noticeable spike at line clear
1800 lbf at 236 psi tank, 145 psi mid, 116 psi chamber,
oddly, this run didnt show a line clearing spike
The 12 long 16 line should be a huge restriction for this
motor, so the absence of a line clearing spike on the last run was odd.
We loaded everything up for a trip out to our remote test
site for some long runs. We welded most
of the remote test site plumbing together out of 2 aluminum pipe, so it should
flow much better than the shop test stand, although there is still a final 4
of 16 flex hose to go to the valve.
Loading the peroxide barrels and big nitrogen six-packs was a good
practice run for our upcoming vehicle testing.
The first run was under about the same conditions as the
last shop run to see the difference the plumbing made:
2000 lbf at 233 psi tank, 159 psi mid, 123 psi chamber, with
a significant line clearing spike.
We then did a long run (11 second burn) at higher pressure,
but unfortunately we lost the tank pressure sensor on the run.
2440 lbf at ??? (about 300?) psi tank, 199 psi mid, 145 psi
chamber, with a line clearing spike of 2550 lbf.
We clearly still have some line loss at the 16 hose or the
2 to 1 adapter preceding it, which accounts for the 15% shortfall from the
A few notes on working with the big engine:
Bolt stretch at the hot flange was significant. We reduced some of our margins in the scale
up from the 5.5 engine, including the bolt clamping force. Final engines will be all welded, but we may
need to drill the bottom flanges out to 5/16
This engine was much easier to preheat than any recent
engine with good performance. There are
a few possible explanations for this starting better than the 1.7 throat small
engine: There may be a scaling factor that favors larger catalyst packs. The 900 cpsi monolith on top may be
decomposing more of the 50% peroxide before the glow plug chamber. For spacing reasons, Russ welded the glow
plug holder at an upwards angle, which might act as a little propellant scoop
during starting. In any case, every run
was easy to start two slugs of propellant to wet everything, then crack the
valve and wait a minute or so for the exhaust to stop smoking and popping and settle
down to a steady invisible flow. We
found that if you start it when the exhaust is clear, but there is still some
popping or condensing liquid in the nozzle, then it will run cloudy for the
first second before clearing completely.
If you wait for it to completely preheat, there wont be any significant
clouds at all.
There is an interesting scaling relationship between engine
size and wall thickness. If you double
the diameter of an engine, the walls need to get twice as thick to maintain the
same level of material stress. However,
for a given chamber pressure, the thermal transfer into the wall stays the
same, so the double thick wall transfers heat out much more slowly. This engine was only just starting to glow
dull red after the 11 second long run.
2500 lbf was not nearly as loud as we expected. Most of us left our hearing protection off
for the second run, and we were only 60 or so away. This is good for manned vehicles.
All the operation at the test site with large quantities of
propellant and nitrogen went very smoothly, which bodes well for the vehicle
testing. We need to reinforce our shed
wall a bit though the first engine run tore some of the trim off when it
This 12 engine, along with four small ones, would give us
plenty of thrust for all the pre-launch-license test flights, but plumbing it
up in the middle without adding another foot of height to the vehicle is
looking like a nightmare. We will be
testing a stepped chamber engine next week, which may give us enough thrust to
fly in the current configuration.
Various small things were also accomplished on the big vehicle:
the master cutoff computer box brackets were made, the external antenna was
mounted, and a custom DB connector panel was made for all the serial ports.
(11 meg video)
I did some more calculations on the Isp for the mixed
monoprop at various chamber pressures at both our normal 5 : 1 and the
stoichemetric 4 : 1 volume ratios:
MR 100 150 200 250 300 350 400 450 500 550
-------- --- --- --- --- --- --- --- --- --- ---
4 : 1 143 154 161 166 170 173 175 178 179 181
5 : 1 135 146 152 157 161 163 166 168 169 171
exp 1.8 2.3 2.7 3.2 3.6 4.0 4.3 4.7 5.0 5.4
Exp is the optimal nozzle expansion ration.
Our measured 145 Isp would have been at about a 175 chamber
pressure with the small nozzle, and at a mixture ratio in between the two
above, so we are actually very close to theoretical performance.
It is clear that there is a big performance jump going from
100 to 150 psi chamber pressure, but the gain gets smaller and smaller as
pressures increase. The hotter mixture ratio is a reasonably constant
bump across the board, but gets a little better at the higher pressures.
Discuss Armadillo updates at the Space Frontier Foundations
Frontier Files Online: www.space-frontier.org/FFO