July 26, 2003 Notes
Big Vehicle Plumbing
We assembled and tested the major fill and drain plumbing
for the big vehicle this week. To be
able to load and drain 800+ gallons of propellant in a short amount of time,
the plumbing is quite large. We are
using 2” cam and groove hose couplings for connections, and we are welding 2”
pipe connections wherever possible, instead of using NPT connections. We machined custom aluminum flanges to
replace the NPT flanges on our KZCO servo valve for the master cutoff, which
allows us to do away with rotating connections there. We still have NPT connections between the stainless steel check
valve and aluminum hardware on the loading side.
Fitting a 1/8” thick O-ring to an 18” diameter groove is a
pain, because the groove isn’t really deep enough to keep it from wanting to
pop out as you move around the circumference.
A thicker O-ring would have made life easier, but then we would have
needed to use a thicker flange plate, which would have added weight and
required longer flange bolts. We used
Teflon thread sealant to stick down the O-ring as we moved it into position,
which worked out well enough.
We got a bit lucky on some of our positioning – it is only
just barely possible to actually assemble all the various bolts and fittings of
the plumbing on the vehicle.
We loaded 30 gallons of water from a transfer tank into the
vehicle tank to check the procedure and look for leaks. Because we don’t have the engine plumbing
assembled we temporarily just put another big hose coupling under the master
cutoff valve for draining. It turns out
that we have a leak around the parachute support post, so we are going to have
to take everything apart to fix it.
This may have happened when we set the vehicle upright, because our
current legs are a few inches too short for everything, and the vehicle hit the
parachute shackle, which may have cracked the weld inside the tank.
http://media.armadilloaerospace.com/2003_07_26/vehicleBase.jpg
http://media.armadilloaerospace.com/2003_07_26/bottomView.jpg
http://media.armadilloaerospace.com/2003_07_26/flange.jpg
http://media.armadilloaerospace.com/2003_07_26/camlock.jpg
More Mixed Monoprop Tests
On Tuesday, we made several modifications to our test
setup. We replaced the ¼” ball valve
and associated plumbing with a ½” ball valve, we plumbed the pack preheat fuel
/ air lines in without a manual valve, and we added a chamber pressure
tap. We also moved the engine mount to
a new beam for mounting on our permanent test stands, in preparation for high
thrust / long duration test burns at our remote site.
The chamber pressure tap has worked perfectly so far. We welded a swagelock fitting onto the
engine, and ran about two feet of fairly small stainless tube from that to a
pressure transducer with a porous metal snubber in front of it. The snubber does act like a low pass filter
on the pressure readings, but we didn’t see any signs of temperature affecting
the readings, so we are happy with it.
All the runs on Tuesday suffered from extreme chuffing due
to the replacement of the ¼” valve with ½” hardware. The chamber pressure measurements showed that we had very little
pressure drop, especially considering that there was still 10’ of –10 hose
between the tank and the valve, meaning that much of the measured pressure drop
was from the hose, rather than the valve / spreading plate / catalyst
pack. A couple of the runs had smooth
enough spots to get reasonable data from, but this was certainly not a goodconfiguration
for the engine.
On Saturday, we swapped out the 1.25” diameter nozzles for
the big 2” diameter nozzle, which would allow much higher thrust, but would
result in a lower chamber pressure, which we hoped would eliminate the chuffing. We tried twice with this, but we flooded out
the catalyst both times, only building a little chamber pressure before
dropping off to nothing but a foaming mess.
At a tank pressure of around 200 psi, the initial inrush of liquid with
no chamber pressure is two to four times as high as when the engine is
operating, depending on how restrictive the plumbing is relative to the
catalyst. This catalyst does not seem to
be able to handle that level of flow. A
properly sized cavitating venturi could prevent the excess initial liquid flow,
or we could try to just crack open the valve until some chamber pressure has
been built.
We added a ¼” restriction before the ½” ball valve to give
more pressure drop for smoothness, and made a few more runs. The low pressure runs were still rougher
than the runs from the 19th, because the single ¼” restriction still
offered less pressure drop than the ¼” valve and fitting combination that we
previously had. The run at high
pressure was smooth, because it had sufficient pressure drop. The measured Isp from that run was 137 s,
consistent with the previous measurement.
I had let it run for a second with the throttle only cracked to allow
some chamber pressure to build, which reduced the total Isp somewhat, and we
are seeing a little bit of channeling around the outside of the catalyst pack,
which also cut a few seconds from it.
We will probably weld in permanent anti-channel rings between the
catalyst slices, as we started doing for our screen based engines.
The last run, again at low pressure, did started chuffing
again, and the catalyst was quenched.
The next test we need to do is determine if the quenching was a result
of the chuffing, or vise versa.
Things are working well enough that I have ordered
sufficient catalyst for four 5.5” diameter catalyst packs, which will be able
to do low altitude flights of the big vehicle, and one 12” diameter pack for
testing the full scale engine. We are
moving from the 200 cells per inch in the current test catalyst to 400 cpi for
all the 5.5” catalysts, and we are going to try an experiment with a single 600
cpi catalyst in the big engine, under the theory that the very tiny cells will
be too small to support liquid tunneling, and we won’t need to segment the
catalyst.
Pulling representative numbers from the three days of
testing:
July 19:
62 lbf from 145 psi tank pressure, 0.120 orifice
113 lbf from 132 psi tank, 1/8” NPT restriction
248 lbf from 328 psi tank
163 lbf from 157 psi tank, ¼” NPT valve and fittings
335 lbf from 366 psi tank
July 22:
278 lbf from 227 psi tank, 179 psi chamber, ½” NPT valve
293 lbf from 226 psi tank, 181 psi chamber
July 26:
199 lbf from 168 psi tank, 138 psi chamber, ¼” NPT restriction
553 lbf from 528 psi tank, 345 psi chamber
199 lbf from 167 psi tank, 142 psi chamber
We also did another experiment with hydrogen based
preheating. Flowing hydrogen / air mixtures
through the engine does indeed auto-ignite after a short period of catalytic heating,
avoiding the need for a manual torch lighting, but once it ignites into a
flame, the flame immediately flashes back into the plumbing above the engine,
putting a lot more heat there than we like.
We will probably try welding a fitting for the hydrogen directly onto
the engine, so that only the forced air flows through the external plumbing.
This will still raise the spreading plate to full flame temperature, but the
plumbing should stay cooler.
News
