July 23 and 27, 2002 Meeting Notes
Another Check Valve
We continued to have odd problems on the test stand, and
they were traced to the brand new check valve falling apart just like the
previous one. They were both of a
similar design, with a Teflon retainer threaded inside after a ball and
spring. We dont know if the peroxide
fumes are somehow weakening the plastic threads, or if there is a vibration
issue causing them to unthread, but we have changed to a different style of
check valve that consists of two metal sections threaded together.
The bench test engine we have been using was the one with
the low compression catalyst pack with slightly fewer screens than our normal
packing for the 2 engines. This always
felt a little marginal, so in an attempt to remove more of the uncertainty from
our biprop testing, we added more screens to bring it up to our normal level
(64 silvers, 50 stainless), and re-pressed it to our standard level of 1500 psi
indicated (about 3000 pounds). I
probably should have just compressed it to fit, because we know that
recompressing a pack after it has been fired will shorten it a lot more than
the original compression, and it did not completely fill the pack. I added a couple more screens, but it was
It was yet another day of uncertain results with ethae
burning, getting two rich looking burns with 0.060 peroxide and 0.55 ethane
jets, but they were not repeatable. The
annular injector ring does not atomize kerosene very well, so we have been
struggling to make the ethane reliable, but we will probably make a new
kerosene injector soon. We replumbed
the kerosene tank so the sight glass doesnt connect at an area of high flow,
and added a separate pressure gauge for that tank. We should have the radiatively cooled nozzle next week, so we
will give the ethane one more try with that, hoping that the hot wall helps
We tested a new retaining plate design on Saturday. Our existing retaining plates have either
been cut from perforated metal plates, or drilled with pattern of 0.25
holes. This gives an open area fraction
of around 45%. As an experiment, we
milled a custom 2 diameter retaining plate that just had a thin outer ring,
and a 0.1 wide cross through the center, for an open area fraction of around
72%. This may improve high flow
characteristics of the engine somewhat.
There was some concern that the screens would bow down into
the larger gaps, but it held up fine for all the test runs. We always end the pack with a stainless
screen, because a silver screen might break wires when compressed against the
All tests were with a 1/4" ball valve instead of a
solenoid, and the peroxide was at 85 degrees F at the start of the day. We run with the normal retaining plate until
the performance was shown to be steady.
The first run was one liter with a 0.120 peroxide jet at 600
psi tank pressure. It was densely
cloudy the entire run. The pack was
still wet from the water flushing on Tuesday, so it might have crudded up a
The next run was an exact repeat that was also cloudy, but a
little less so.
The next run dropped the peroxide jet to 0.090, but it was
We were getting concerned about this, but we had noticed
with the rotor engines that sometimes a poor pack just needed to be run
continuously to break it back in, and no amount of short pulses would do it.
The next run was a 0.070 peroxide jet at only 300 psi, and
it soon started running perfectly clear and smooth at 25 pounds of thrust.
The next run went back to 0.100 peroxide jet, but stayed at
300 psi. It also ran perfectly clear
and smooth at 46 pounds of thrust.
The next run was 0.100 peroxide at 600 psi, which made a
perfectly clear and smooth 85 pounds of thrust.
The next run was 0.120 peroxide at 600 psi, which made a
perfectly clear and smooth 108 pounds of thrust.
We then switched to the new retaining plate and ran it again,
getting a perfect 114 pounds of thrust.
However, on looking back at the data, the pressure was 15 psi higher, so
the increase may not have been solely due to the retaining plate open area.
All of the perfect runs were almost perfectly flat thrust
curves, even though the tank pressure would decay 30 or 40 psi over the coarse
of the runs. This implies that the jet
was still providing enough pressure drop to make the flow proportional to the
square root of the feed pressure.
We replaced the jet fitting with a 1/4" to 1/8 NPT
nipple directly from the ball valve to the engine. The fitting had an ID of 0.180, so it has over twice the cross
sectional area of the jet. I was
expecting the run to be rough, but it was perfectly clear and smooth. The load cell we were using only read up to
135 pounds, and it clamped there for the entire run. Based on a burn time of 2.3 seconds for a liter of peroxide, it
was likely around 154 pounds of thrust.
We worked on a bunch of other things for a while, and
eventually got a bigger load cell mounted on the test stand. When we run the engine again, it was
full-cloudy again, blowing all the peroxide out in a very short amount of
time. We warmed the engine more slowly
with another run, then we got another perfectly smooth and clear run of the
same duration as the last test in the first run, but the load cell only registered
98 lbf, so I am pretty sure the calibration or test stand geometry is off. That is also in keeping with our
unexpectedly high acceleration from the large motor, which may have been
underestimated in thrust due to test stand geometry.
This is around 50lbf of peroxide per square inch of catalyst
pack, which is much higher than we normally run. Another point of note is that the catalyst pack was loose enough
in the engine that you could hear it when you shook the engine. This may have contributed to the early wet
runs, but it does not seem to impact the ability to get smooth, high thrust
The larger open area test plate does seem to have some
benefit, so we will probably make our future plates in that manner. This test plate was made out of brass, so we
cant use it for biprop testing, but when we get our CNC mill installed, we
will make some stainless ones.
We are probably going to try minimizing the water soaking of
our catalyst packs in hopes of avoiding the long break ins like we had today. On the test stand, we can disconnect the
plumbing after the valve and continue to rinse the rest of the system, but that
isnt an option on the vehicles. When
we do flush with water, we will try to let the entire nitrogen tank blow down
through the engines, rather than opening up the main purge valve.
Tube vehicle checkout
We carried the tube out to water test it and hot fire all
the engines, but not lift off. After the
long break in period with the test stand motor, we were expecting to have to
run these for quite a while, but they cleared rapidly, and everything seems to
still be in good order.
The remote pilot program crashed during our testing, but it
seems to be repeatable, so I should be able to fix it tomorrow.
We have enough connectors on the electronics box now that it
is getting to be a hassle dealing with all the discrete cables: attitude
engines (all four in one connector), main engine, GPS, altimeter, parachute
signal, pressure, temperature. I am
going to make a single 37 pin CPC connector that carries all the signals from
the box to the vehicles. The individual
sensors and actuators will still have separate connectors, which will be
connected to a breakout connector that stays permanently mounted to the
vehicle. This adds an extra set of
contacts to the electronics, but the system operability will be improved.