June 13, 2004 notes
Streamlined hovers and landings
We completed the streamlined subscale vehicle and conducted
several hover tests this week. Total
vehicle weight is 310 pounds dry.
We did some drop tests from the hoist to evaluate the shock
absorber landing gear, and they did a very good job, even when the vehicle was
swung before release. Without any
bounce, the vehicle just settles down, even at slight angles.
It has been raining a lot, and it wouldnt let up for us at
all on Tuesday, but we really wanted to test things out, so we put a big plastic
bag over the rocket nose to give some extra cover to the electronics, and went
ahead and gave it a try. The flight
control software had been modified to predict acceleration some time in the
future to compensate for the time lag between moving the valve and engine
chamber pressure changing, which should smooth out the up/down oscillations in
hover and landing modes.
We stood the vehicle up on the big foam blocks so we didnt
need to trust it to come down straight, but this put them too close to the engine
on the narrow vehicle, and we burned off a lot of the aluminum covering we
added. We are probably going to get
some stainless foil next time.
This engine didnt warm up as nicely as the engine on the
lander, which we attribute to the lack of a compressed hot pack. After it finally got warm, the hover looked
very good, but it ran out of propellant before it landed, which made us very
happy we had done a block launch instead of a ground launch. We did another run with extra propellant,
and everything went normally. The
acceleration prediction seemed to be working very well.
At hover, the chamber pressure was only 100 psi, so on a
boosted hop with even a 0.5G stabilize level, we are looking at a 50 psi
chamber pressure, which is certainly going to separate inside the nozzle. We seem to have pretty good luck with even
separations, but we have certainly seen some single-wall attachment
separations, which I have been very concerned about disrupting flights. We decided to cut off some of the nozzle,
taking it down to only about a 1.8x expansion ratio. This will be underexpanded on boost, and still rather
overexpanded during coast, but it will hopefully be a reasonable compromise.
We also cut the engine open to make a compressed hot pack
for better warmup. There was one burned
spot in the catalyst retainer, but overall it didnt look too bad. We replaced the catalyst anyway, and added
one of the thicker water jet cut retaining plates so we could put 3000 psi
gauge pressure on the 7 diameter pack.
With the extra height of the new retaining plate, we werent quite able
to get all 1000 grams of catalyst in.
When we were setting up for a ground liftoff test, we bent
one of the shock absorber rods while lifting it off the cradle. It was just carelessness in this case, but
it would be unavoidable on a big vehicle to lean on the shocks, so we are going
to have to make removable covers or some other kind of removable support when
we move to shocks there. We bent the
rod back into shape, but it definitely has an uneven landing to it now.
The compressed engine hot pack did make the engine warm up
steadily without needing any throttle adjustment, but on throttle up it did
show a stream of clouds in the exhaust, meaning that the cold pack isnt in
great shape. We dont have any unused
5.5 catalyst monoliths, but we have 7 chambers coming, so we will eventually
make a completely fresh engine just for this vehicle.
On liftoff, the valve insulating cover ripped off right at
throttle up. We arent exactly sure if
this is due to incremental wear on the insulation allowing the epoxy to come
apart, the cut down expansion cone putting more pressure on the side, possibly
in conjunction with the pressure spike this flight had at throttle up, or just
the proximity to the ground in this aerodynamic configuration. We are making replacement covers in metal,
and screwing them to the bulkhead plate.
Contrast the altitude hold and landing with previous hover
tests. I should be able to get the
horizontal position hold tighter, but we are very happy with this
behavior. I am concerned that the
visible exhaust cloud seems to be coming off at a fairly large angle, but it
seems to have plenty of control authority.
If it doesnt rain any more, we are going to try a boosted hop with this
vehicle on Tuesday.
I have the software set up now to take differential GPS
corrections over our normal telemetry stream, but we were getting such good GPS
signals today that I wasnt able to tell if it got rid of the discontinuities
in the GPS stream. Having the antenna
really at the top of the vehicle, with a nice metal ground plane below it, may
well be giving us a lot better GPS reception than in the previous vehicle,
which had some metal bars above the antenna, and a composite base.
We also set up the remote battery boxes in the big vehicle,
and rebuilt the 2 valves that were allowing propellant to leak into the engine
during pressurization. We are still
afraid that we might have killed this 12 engine due to the leaking during our
last test. Trickle flows get the engine
started without any ignition source at all, because the methanol decomposes on
the catalyst into hydrogen and formaldehyde, and the hydrogen will
catalytically burn with oxygen even at room temperature, eventually raising the
temperature enough to start free-space combustion. Unfortunately, this allows burning to take place in the cold
catalyst pack, because the flow isnt high enough to keep it down at the
flameholder where we want it. This
starts locally killing the catalyst.
The lesson is to not live with leaky valves.