June 11, 2008 notes:
We are working hard on the Rocket Racer program and making
good progress, but the Rocket Racing League has requested that we allow them to
manage the public dissemination of information about the project, so I wont be
discussing it here. Since that has been
a large part of our work lately, this update is a bit light.
We are finally completely moved in at our new location. Getting our three-phase power hooked up took
a lot longer than expected, but we are finally fully operational parts have
been milled, turned, and welded. It will
be another month or two until everything finds its proper place here, but
things should start moving along at an accelerated pace now.
We have run over ten thousand pounds of propellant through
the test sled this month, and I think everything is at essentially final form.
I had known for a while that the alignment of the angle
table we use for drilling the injector elements wasnt quite right we had
noticed that the elements werent perfectly impinging, but things still seemed
to be working ok. When I finally got
around to dialing it in very precisely and made another injector with it, we
got a pleasant surprise performance jumped significantly, and interestingly,
the engine actually ran significantly cooler!
Previous engines must have had stray oxygen burning with the film
cooling fuel at the wall, while the new engine gets all the oxygen properly
burned with fuel from the elements before it has a chance to get to the film
We accidentally did an interesting test with the new
engine. We were trying out different
regulators for the startup / shutdown purge gas, but we wound up removing the
regulator on the fuel side without changing the code to only pulse the purge
solenoid, resulting in a continuous, large amount of helium being dumped into
the fuel manifold while the engine was running.
Surprisingly, it still ran perfectly smoothly, but thrust was down a
bit, and the exhaust bell started glowing red hot after ten seconds of
firing. We ran it a few times, and
eventually eroded a little bit of metal out of the exhaust bell when we were
seeing if it was a steady-state red hot or not (not quite). We were very puzzled why the engine had just
changed performance and heating dramatically until we realized the issue with
the purge solenoid. Upon analysis, we
found that the helium had displaced over a third of the fuel, resulting in the
engine running right around stoichemetric mixture
ratio with reduced film cooling. The
fact that the engine could almost run steady state under those conditions was
extremely encouraging. We could
certainly lean the properly fueled engine out a fair amount with no
consequences at all.
We had been chasing some igniter problems for a while now,
where the igniter would light properly, but abruptly flame out after a few hundred
milliseconds when operated at feed pressures below 200 psi. We had never run into that before, because
our vehicles never had to do restarts after the initial pressurization. Because we eventually intend to shutdown the
engines and relight for landing after high altitude VTVL flights, this was
obviously something that had to be solved eventually. Our ignition interlocks handled the flameouts
properly (important point: dont assume that an igniter will stay lit, a single
point check for pressure is not sufficient!) every time, but it was
frustrating. It looked like it was
related to the flow inrush when the main propellant valves started opening, so
we spend several tests moving around the igniter propellant pickup points and
adding accumulators, but we eventually fixed it by changing the igniter injection
geometry. We now light reliably down to
100 psi (and probably lower), but our spark plug is taking a lot more heat with
the new geometry, so the last thing we are probably going to change is to move that
a little farther away.
At this point, I can say that the decision to go to the film
cooled engines was absolutely the right thing to do. We are improved on every metric from the
cooled graphite engines of last year.
One other interesting point that all the restarts in our
testing lately showed up: normally, we
dont really have to worry about our lox heating up too much in the test stand
tanks unless we take twenty or thirty minutes to fix something, but the
situation changes a lot if you pause when a spherical tank is nearly
empty. The high surface area to volume
ratio of a shallow puddle at the bottom of a sphere sucks heat up surprisingly
quickly. We had a case where we were
paused before the last ten second run on a tank, and the lox managed to heat up
enough to self-pressurize beyond the 125 psi that the tanks were sitting at
from the end of the blowdown run. This generally isnt a real problem, because
the mixture ratio just gets richer as the lox gets fluffy, but we try to
We tried to do another long, high thrust methane run this month,
but it didnt work out well. There were
several loud bangs as it started up, then the engine
chamber peeled itself off the injector.
Upon inspection, it turned out that the weld penetration wasnt very
good, and it was a purely pressure related failure, rather than a burn through,
but the engine clearly wasnt running right.
The methane engine that fired ok two months ago had just a single ring
of unlike doublets, but the last two ones we have had problems with have been
injectors with two rings of split-triplet elements. Our current belief is that the separate ring
manifolds are causing problems during startup as the methane chills first the
center, then the middle, then finally the outer ring. The lox travels around in one manifold on the
top, feeding all the elements at about the same time. This means that the inner ring starts at
about the right O/F ratio, but the outer ring is extremely lean, and the film
cooling holes just get gas for some time.
The outer ring may well have been too lean to ignite cleanly and
promptly, resulting in the bangs.
It is so nice to be testing at our shop again we were able
to clean everything up from the somewhat destructive failed methane engine test
and perform two more (successful
) alcohol test firings later that same day.
We currently dump methane from the plumbing in front of the main
fuel valve to make sure we have liquid up to that point, but we are going to
try dumping some methane through the engine next time to pre-chill the
manifolds before starting. We may yet go
back to a single ring of elements for methane, but we had experienced injector
face heating problems with that, and also generally lower performance. It is possible that with a 310 alloy face,
possibly with a coating, and the proper alignment on the holes, it may not be a