March 1, 2006 notes
We tried to hover the new vehicle,
but it didnt work out. We got the
vehicle up on the stands, and everything went smoothly with the flight
checklist, but when I throttled up the vehicle didnt lift off. I eventually kicked the gimbal
around with the joystick to knock it off the stands to see how it behaved
swinging, but it never built up any more thrust. From all indications, it would have flown
properly if it had more chamber pressure.
This engine had mode over 1100 lbf
on the test stand previously, but when we tested it again it was down to 800 lbf. However, even
at 800, it should have picked the vehicle up.
An additional loss of thrust is probably explained by higher
lox temperature / lower density due to greater heat transfer into shallowly
loaded large tanks versus highly filled small spheres on the test stand, and
the longer time between closing off the tanks and firing the engine on the
elevated vehicle versus the test stand.
We are considering adding insulation to the lox tank to help this, but
it shouldnt be necessary if we arent chasing absolute maximum
performance. We got some spray
polyurethane foam to evaluate, but we might be better off with something like
mineral wool that would be non-flamable.
For testing this vehicle I purchased a set of race car
scales, which lets us tell exactly how much lox is loading and boiling off
during the filling process. It also let
us know that the previous weight we had for the vehicle was off a bit due to
maxing out the old floor scales we had the actual vehicle dry weight is 600
pounds even. At full load it will hold 1500
pounds of propellant, for a mass ratio of 3.5.
Using real bearings and shafts for the gimbal
hinge points reduced the slop in the system considerably. Phil thinks that much of the remaining slop
in the actuator can also be corrected by shimming some internal gears, but it
is good enough for now.
The engine that was on the vehicle (4 ID regen chamber with radial injector, 2.75 throat graphite
nozzle) is definitely damaged internally.
We havent cut it apart yet, but it appears to have a failed weld. This engine had the top popped off with the
old igniter and repaired, as well as a couple configuration changes, so we
arent shocked that it isnt healthy.
I had started milling a 6 ID regen
chamber for a new motor, but I wound up messing up the first one, and due to
some design considerations for what we thought the centennial challenge prize
was going to require, we decided to change designs and go with an uncooled chamber again.
The graphite nozzles have been holding up impressively
well. The zirconium coatings we tried
did not last long, but a silicon carbide based coating seem to do much
better. Our current testing is with a 12
long 6 ID / 8 OD phenolic tube over a 3 throat
graphite nozzle, but we are having an experimental chamber fabricated that will
be solid graphite with a carbon fiber reinforcement layer. We have reasonable expectations that such a
chamber will be able to run for many minutes of continuous firing.
We wanted to experiment with axial propellant injection now
that the head ends are getting to decent proportions. Radial injection was nice for the smaller
engines, but the fabrication is less repeatable with the runout
on tubing (and in my mill fourth axis chuck one of these days I will fix
), and we have a definite issue with stacking radial lox injection points
over the radial fuel injection points as the engines scale up. Early French engines and the OTRAG engines
use alternate posts of fuel and oxidizer injectors around the circumference of
the engine, which apparently scales well to larger engines, but I cant
conveniently machine those with my current setup.
The new injector plates have a central igniter port, and a
single ring manifold each of fuel and oxidizer with 60 degree unlike impinging
injectors. The current one has 80 holes
each, all 1/16 diameter. This flows
slightly fuel rich with ethanol, but it should improve chamber life. This should be the rough size that we use for
quite some time somewhat over 3000 lbf. One engine on the current
vehicle, and four engines on the upcoming 65 differentially throttled vehicle.
Working out the hole drilling process was important. I drill from the chamber side, first milling
a land at the 30 degree half angle, then spot all the holes with a spot drill,
then drill through in two pecks (about 0.25 total depth for a 4:1 L:D ratio), then go back and mill down another 0.06 to clean
off the enlarged spot drill holes and leave a clean injector exit. The holes on the inlet side have the final
chips cleaned off manually.
We made the torch igniter / engine mount so that it screws
onto a -16 AN fitting, making that section portable between different injector
It takes about six hours of milling to make one of these
injectors, so we can test a lot of variables quickly. Version one had 90 degree injection angles
and didnt have the spot holes cleaned off, which resulted in an obviously messy
injection pattern with water. We fired
it and it worked, but performance wasnt great.
Version two has the 60 degree angle elements with the ends closer
together and other changes necessary to allow this, as well as improved pockets
for assembly on the top side and chamber centering, and a smaller flange bolt
circle with more holes. This injector
burned through in operation, but it logged by far the best Isp we have seen (details after we have a perfectly
clean run of data to back this up).
Version three will add tapered manifold inserts to keep the propellant
velocity uniformly higher in the rings for better face cooling, and make more
changes to ease assembly of the main propellant inlets.
Cold flow tests on version two looked a lot better, but
there were still a couple small stray sprays that werent axial with the rest
of them. I may try reaming the holes to
final diameter on a later version, or drilling with tiny end mills, but for now
we just want to replicate version two with better cooling.
We blew out the graphite gasket between the nozzle and the phenolic chamber while testing version one. For version two, we took a belt-and-suspenders
approach version two upped the bolt count on the flange and brought the holes
closer in so it could take more torque without bending, and we machined a
pocket in the phenolic tube to capture the gasket and
center the nozzle.
The phenolic tube is holding up
better than we expected. The 2:1 diameter
contraction ratio makes a huge difference in ablation rate compared to the
chambers on the X-Prize Cup vehicle, which only had a 1.25 or so diameter
contraction, and ablated rapidly.
Between the graphite nozzle and the aluminum retaining
flange we have been using a half inch thick ring of machinable
ceramic, but I am sick of it, because the rings are so brittle that they break
every time we torque the bolts down. It
is just a compression spacer, so it doesnt really hurt anything, but I want a
better solution. We added a pocket to
retain the chunks and sprayed some adhesive on the back so it all stays
together, but it still sucks. We will
probably try a phenolic spacer or a solid stainless steel
flange. Some brake pad like material
would probably be a good insulating spacer.
Anyone have suggestions?
Clamping graphite nozzles under phenolic
tubes probably isnt going to work for long runs, because the phenolic is ablating at the contact point of the gasket,
leaving a surface that will probably leak if we run it long enough. We hope the solid graphite engines work out,
but if we wind up using an ablative chamber with a graphite throat we will
probably have to make a tapered joint between them and slide the nozzle in from
the top, so it continuously pushes down into the charring ablative.
As the engines have been getting bigger, we have been
chewing through our firebrick blast deflectors faster. On the last refurbish we added a graphite
plate on top of the bricks, and it doesnt seem to have suffered at all after a
ten second burn.
1500 lbf (running at reduced feed
After the injector base burned through, it rapidly burned
through the top closures and sprayed lox above the engine, charring a lot of
wiring and setting the computer vibration insulation foam on fire. Putting that out with the fire hose resulted
in us getting some water inside the electronics box, which required us to
actually replace a PC104 connector due to corrosion. At one point the box was sealed reasonably
well, but the gasket needs to be replaced, and we should conformal coat
The lunar lander
centennial challenge is our top priority this year unless something else pops
up. We had a commercial opportunity that
was exciting, but it seems to have fallen through.
Im not thrilled about landing on inclined, boulder strewn
fields, but the payload and delta-V requirement are easier than we
expected. Having two levels and
consolation prizes is a good thing.
As soon as we can show that the new engines can make two 90
second burns, the current vehicle should have level one in the bag. We will need software changes and a remote
video system, but no other significant modifications. To take the big level two prize we will need
a completely different landing gear arrangement, and the total performance may
be pushing it a bit. If our new engine Isp is as good as it briefly
looked, we may be able to modify this vehicle for level two, but we are
expecting to have to use the upcoming 65 diameter vehicle, which will have a
better mass ratio.
It is unfortunate that the prizes can only be claimed at the
X-Prize Cup, because that will encourage us to sit on the vehicles after they
have been proven out, rather than flying them hard and potentially crashing