September 19, 2004 notes
Our latest stab-in-the-dark on the 7 mixed monoprop engine
was to try welding a 1.7 throat nozzle on instead of a 2.2 throat, but it was
still rough at about the same throttle level.
Our next stab-in-the-dark is trying a hot pack made out of the old
catalyst bale that we used before we moved to the ring catalyst. We will be testing that on Tuesday. We arent giving up on the combination yet,
because we have over $50,000 worth of mixed-monoprop vehicle stuff on hand in
catalyst, nozzles, support plates, and tanks (the tanks and nose cones to
replace the 48 diameter vehicle have arrived).
However, we have been working on the preburner lox /
methanol engine, and it has been going easier than expected.
We implemented out new burner scheme with a long enough
central pipe for the gox and ethanol to do most of the burning before it gets a
chance to mix with the lox flowing around the tube, and it seems to work
well. We estimated the tube length
necessary by starting with a very long tube, firing the burner, then just
cutting it off at the point where it seemed to reach a constant
temperature. This should be
conservative, because at operating pressure it should combust faster.
The vaporizer would work fine being just a long jump
enclosing the burner, but mounting it would be a bit difficult, so we welded it
to one of our monoprop tops, which allows us to feed fuel through a tube
straight down the centerline. The
vaporizer tube is off center, so the hot core doesnt directly impinge on the
fuel pipe. We have three sensor taps on
the chamber top: vaporizer temperature, vaporizer pressure, and chamber
pressure, which is a tube going down through the vaporizer top and closure
plate to sense the chamber pressure. We
need to sense like that, because the real combustion chamber will have a
cooling jacket, making it difficult to add a pressure tap.
The burner jets are set at 0.100 for gox and 0.050 for
ethane, and both regulators are set at 160 psi.
The vaporizer exit hole is 0.9 ID, with a 0.5 OD tube in
the middle for the methanol. This is
0.44 square inches for the gox, or the equivalent to a 0.75 diameter hole. We initially had a 0.9 ID tube extending
from the vaporizer for about an inch, which would serve as a tight mixing area
before exhausting into the main chamber, but that seemed to burn off almost
immediately when we added the chamber, and doesnt seem to be necessary.
The attachment to the load cell and test stand is somewhat
rickety, which gave us some vibration noise in the load cell measurements.
There is a lot to hook up for this engine: lox feed,
methanol feed, gox feed, ethane feed, nitrogen purge, spark plug, ignition
power, and three solenoids.
Our startup procedure is to begin the lox flow to let the
feed line and burner get chilled down, then, when the vaporizer thermocouple
reaches about 0C, start the burner, then rapidly crank up the lox flow. We can play with the lox flow as long as we
want, increasing the flow decreases the vaporizer temperature and increases the
vaproizer pressure. This does keep us
from having precise run to run vaporizer pressures, because we cant just open
the lox valve the entire way without it dropping below freezing coming out of
the vaporizer. Just the vaporizer made
about 50 pounds of thrust at 94 psig pressure.
We will be going to computer control of the burner oxygen / fuel pressure
to allow us to heat more lox.
It does look like the gox coming out of the vaporizer still has
a significant variation in temperature, because there was a visible flame in
the center even when the thermocouple was only reading 200 C. The swirl around the chamber top probably
pushes the cooler / denser gox to the outside.
A turbulent static mixer would probably do a great job of evening it out,
but it would have to be spaced pretty far away from the burner to keep it from
being melted on startup.
We triggered the methanol injection without any chamber
below it to watch the spray combustion pattern:
Autoignition was instant in the hot gox.
We then bolted on an uncooled chamber made of thick stainless
steel with a 2 diameter throat.
We were running 100 psi methanol through a big solenoid,
which flowed about 0.4 pounds per second to atmospheric pressure.
When just the vaporizer was running, there is only a couple
psi of pressure in the chamber, but when the methanol was triggered, the
chamber pressure jumped to 30 psig and it made about 125 pounds of thrust. The vaporizer pressure was 66 psig, and it
was clearly sonic chocked, because that pressure didnt budge when the methanol
was pulsed on or off.
There were a couple bright flashes as the mixing tube and
part of the chamber pressure tap burned off and flew out the nozzle, but the
pressures were dead smooth through the entire run. The heating behavior was interesting it took longer than I
expected for the chamber to start glowing, sitting there for a few seconds
without any visible change, then very rapidly running up to orange hot while I
shut off the methanol. The chamber
appeared to be almost the exact same temperature from only an inch or two below
the flange, all the way to the bottom.
This implies that it was getting very rapid combustion, and the chamber
didnt need to be nearly that long.
When you have a sonic choke, you can determine the mass flow
through it based only on the pressure, temperature, and gas composition. This page has a chart for various gasses:
(note that pressures are absolute, not gauge)
Based on this, we were flowing about 45 pounds of oxygen a
minute, or 0.75 pounds a second. With a
70 psi differential, the methanol was probably flowing around 0.33 pounds a
second. Ideal lox / methanol ratio is 1.25 : 1, so we were quite lean.
On Saturday, we installed and calibrated a turbine flow
meter for the methanol, so we could get better data. I used a 5 50 gpm model, so we needed to push the pressure up
quite a bit to get it in range. For my
future reference: low calibration 1.3v = 6.1 gpm, high calibration 3.6v = 26
Saturday firing #1:
150 psi methanol
3.91 gpm methanol, which is outside the calibration range,
and possibly not to be trusted.
88 psig / 102 psia vaporizer pressure
68 psig / 84 psia chamber pressure
Not quite critical flow from the vaporizer, the vaporizer
pressure changed when the methanol started and stopped. However, it still had perfectly smooth
combustion, so decoupling the vaporizer from the combustion chamber may not
actually be necessary.
Saturday firing #2:
To increase the vaporizer to chamber pressure ratio, we welded
a reducer onto the vaporizer exit, taking it from 0.9 to 0.79 diameter,
reducing the gox flow area to 0.3 square inches, or an equivalent diameter of
0.62. We welded an extension onto the
methanol tube so it protruded well below the exit plane of the choke plate to
guarantee that the methanol injection wouldnt have any impact on the vaporizer
pressure. We left it about an inch
over, because we were curious how it would effect the combustion pattern.
445 psi methanol
8.12 gpm methanol
78 psig / 92 psia vaporizer pressure
66 psig / 80 psia chamber pressure
This was a little bit odd, because the flow was critical
from the vaporizer, not changing at all when the chamber pressure pulsed from 2
psi to 66 psi on methanol injection, but the pressure ratio was even lower than
the previous test due to the increased methanol going into the chamber. There are some details about thin plate
sonic chokes and pressure recovery that may explain the discrepancy.
The heating pattern was much more splotchy than the previous
runs, which stands to reason. We want
to get the methanol injection point as soon after the sonic throat as possible
to give it the highest speed gasses to tear the low velocity stream apart. The next time we open up the engine we will
cut the methanol tube off flush with the bottom of the choke plate.
This run was too rich, so we dropped the methanol pressure,
which should also drop the vaporizer to chamber pressure ratio.
Saturday firing #3:
300 psi methanol
6.43 gpm methanol @ 6.63 lb / gallon = 0.71 lb / sec
92 psig / 106 psia vaporizer pressure = 45 lb / min = 0.75
lb / sec
73 psig / 87 psia chamber pressure
250 lbf thrust
This also had critical flow from the vaporizer. This is slightly rich, so this area ratio of
0.3 effective diameter at the vaporizer to 2 diameter at the chamber throat
does have a little bit of margin, but it would probably be good to give it a
When we shut off the burner at the end of a run, the top
chamber frosts over by the time we shut of the lox, but the chamber retains
enough heat to still be glowing:
We had the inner section of our cooled chamber hardcoat
anodized, so it is now ready to use. We
expect that it will be able to handle low pressure runs like this, but probably
not much more, because it has a full inch of aluminum between the cooling
jacket and the throat. The throat will
probably start to melt out when we push the pressures and heat transfers
up. We need to make a saddle style milled
slot cooling jacket, like the XCOR and SPL designs.
All in all, this seems to be going very quickly. All credit to Charles Pooley for
evangelizing the lox preburner concept to me at Space Access this year. We are probably going to try doubling the
chamber pressure and getting regenerative cooling working at this size, and
possibly building a vehicle of the rough size of our 24 diameter test vehicle
using a lox / methanol engine.
We might skip the 4 throat size and go straight to an 8
throat after that, which is a 15,000 20,000 lbf engine.