October 17 notes:
No update last week because most of us were at the Space
Frontier Foundation 04 conference. We
debuted a new Armadillo video, which is now linked off of the Armadillo home
page. The high point was a long talk
with Burt Rutan the day we left
The new vehicle base is almost completely assembled. We still have to add braces between the
shock mounts, fabricate insulated housings for the vane actuator motors, and do
final welds, but it should be ready for initial operational testing soon.
(jet vanes not installed)
The new differential pressure transducer finally works, so
we will have a fuel gauge for the next flight.
The total set of plumbing on the vehicle is:
On the gas tube:
loading vacuum port / manual pressure vent
piloted solenoid for drogue ejection / computer pressure
low pressure side of differential pressure transducer
possibly the gas source for engine bubbling if we need it
On the liquid tube:
engine valve o-ringed slide seal
propellant loading check valve
propellant drain valve
Directly on the manway cover:
high pressure side of the differential pressure transducer
(not on the liquid tube to avoid velocity effects)
On top flange:
pressure relief valve
LOX Engine Work
We finished the multi part welded cooled engine assembly,
which is getting hardcoated. The CNC
mill I bought two years ago had a fourth axis rotary chuck, but we had pulled
it off the table to free up space early on.
We finally hefted it back up to mill channel slots on the motor, but it
turned out to have a huge 0.090 runout.
I couldnt figure out any way to adjust it, so I wound up actually
milling the jaws until they were concentric, which seemed to work out ok. I designed for 20 ft/sec coolant flow at 500
lbf, but the next one I make I am going to use less channels and make them
deeper so any remaining runout error in depth is less critical.
We tried putting a turbine flow meter on our LOX line, but
it seems to freeze up basically as soon as the line gets chilled down. We may try drawing a vacuum on it to see if
the problem is atmospheric moisture.
We did some firings with the small throat uncooled chamber,
and our theory about subsonic exit from the lox vaporizer does seem to be
correct when the chamber pressure was high enough to keep the vaporizer exit flow
subsonic, combustion seemed much better, melting the throat out in
seconds. As the throat got bigger,
chamber pressure decreased and the vaporizer exit went sonic again. Interestingly, even with the chamber
pressure dropping significantly as the throat widened, thrust stayed almost
perfectly constant due to the flow relationship.
All of our lox engine firings so far have used separate
cylinders of compressed oxygen and ethane to fuel the vaporizer. This was convenient and safe for experimentation,
but we certainly wouldnt want to have two additional consumables on an
The first step was to T off the lox line before the main
valve and make a coil of thin tubing to let atmospheric heat vaporize a small
amount of lox for the burner. This
The next step is to construct a burner that runs on liquid methanol
instead of gaseous ethane. The smallest
spray nozzle I had on hand was a 0.040 hole Bete P nozzle, which flows a
good deal more than we need for the burner, at least with full operating
pressure behind it. We made a number of
burner tube firings with the methanol burner, but we had to take the methanol
pressure all the way down to 50 psi and increase the gox jet size before we got
something that looked near the right mixture ratio. More work is needed. Our concentric
tube burner arrangement also left a lot of trapped methanol volume behind the
spray nozzle on shutdown, so we need to either add a nitrogen purge or redesign
to minimize the volume.