June 4, 2007 notes:
I wish I had a bit more time to go on about these
significant events, but I am crazy busy with work right now. Matt has been in a similar situation, so the
media coverage isn’t as rich as it would otherwise have been.
192 Second hover
We did a full propellant load flight at our local test site,
and finally, everything worked perfectly for it. It went into auto-land after 190 seconds of
hovering, and shutdown was completely nominal.
Everything was fine with the vehicle.
The fuel was almost depleted, but there was quite a bit of lox
remaining, so we are making a new injector that reduces the fuel flow by
cutting the film cooling in half. This
will improve both Isp and
equalize propellant distribution, but the engine will run hotter. If we have a problem with this, I will put all
the film cooling back and just modify the size of the main fuel element
holes. We are still planning on moving
to a lightweight honeycomb electronics box to save weight, and we are carrying
quite a bit more than the required 25kg of payload after we charge back our
safety systems and “gold box” AV equipment, so in final flight form we should
have over 210 seconds of flight time, which is plenty of margin.
There is a definite slosh wobble, but it never seems to
bother the vehicle too much. We have
slosh baffles in the new modules to reduce it.
http://media.armadilloaerospace.com/2007_05_12/192secondHover.mpg
The longest flight of the DC-X was 142 seconds. A tethered hover doesn’t really qualify as a “flight”
in my book, but we will do the free flights soon enough. We already have the fastest demonstrated turn
around time between flights by a huge margin.
I don’t think we will ever fly Pixel to 3140 meters to exceed their
altitude record, but we might (if a customer wants that flight profile, we can
oblige).
It takes a few qualifiers, like “The longest flight of a terrestrial
rocket powered VTVL”, but we are doing things now that nobody has ever done before. It still isn’t at the point where what we are
doing has a clear value to enough people to make an obvious business case, but
that is coming soon.
Full LLC1 flight
One June 2, we conducted a complete LLC 1 operational
profile at the Oklahoma Spaceport.
Everything went great. Representatives
from AST and the X-Prize Cup were present.
This was the first flight under experimental permit rules from a
licensed spaceport. Both legs of the
flight landed within a meter of the pad center, and our operation time was only
an hour and a half.
http://media.armadilloaerospace.com/2007_06_03/LLC1demo.mpg
Oklahoma
representative Jack Bonny took a bunch of nice pictures during the operation:
http://media.armadilloaerospace.com/2007_06_03/PICT0081.jpg
http://media.armadilloaerospace.com/2007_06_03/PICT0083.jpg
http://media.armadilloaerospace.com/2007_06_03/PICT0085.jpg
http://media.armadilloaerospace.com/2007_06_03/PICT0088.jpg
http://media.armadilloaerospace.com/2007_06_03/PICT0096.jpg
http://media.armadilloaerospace.com/2007_06_03/PICT0102.jpg
http://media.armadilloaerospace.com/2007_06_03/PICT0107.jpg
http://media.armadilloaerospace.com/2007_06_03/PICT0111.jpg
http://media.armadilloaerospace.com/2007_06_03/PICT0115.jpg
I want to extend a big thank-you to the OSIDA guys for truly
going above and beyond what could be expected to help us get this done. It turned out that our initial plan for the
launch and landing pads wasn’t going to work out, and Neil spent the entire
week up there with them making things happen.
The lovely range with two concrete pads was an overgrown field the week
before. It looks like AST is going to
allow us to do some higher altitude flights there, so we will likely be making
quite a few trips up there in the next year.
Space shots are still going to have to be from Spaceport America in New Mexico,
but we will be able to hit our max-Q conditions in Oklahoma which will save us a whole lot of
driving.
The "engine cough" that you can see at one point in
the video was actually seen a few times in the flights. It happens when the vehicle changes from a
hover to a 4 m/s descent rate, and drops the throttle almost to idle to get to
that rate as fast as possible. Below 10 meters, the descent rate is only
1 m/s, so we never saw that in the tethered tests. I’m not exactly sure
why it seems to be more pronounced than it was on the flights at the X-Prize
Cup, but the change to a regeneratively cooled engine
certainly effects transient behavior. It
doesn't seem to hurt anything, but I am going to probably clamp the maximum
desired acceleration a bit lower in the future.
As always, we have several things that we are going to
improve before the next flight. Among
other things, I am going to add "view switching" on the piloting
software so I can zoom in for precision landing. To fit the entire 100
meter track on screen doesn't leave a lot of pixels for final positioning in
the current arrangement. Our “discontinuity-tracked-integrated-GPS-velocity”
positioning approach is working perfectly, so with a zoomed in window and some
improvements in our on-board cameras, I think we have a decent chance of having
reproducible landing accuracy in the couple-feet range.
If it weren’t for the X-Prize Cup doing the management of
the NASA prize, we would have won it last weekend. I understand the reasoning behind tying it to
an event to help promote the industry as a whole and provide more opportunities
for other teams to catch up with the front runner, but as the front runner, I
would rather have the check…
I have some holes in my weekend schedule coming up, so we
may not be able to get as many tests done in the next month, but the plan is to
do back-to-back 180+ second runs with the tweaked injector and final weight
cuts locally, then go back to Oklahoma for a repeat of this exercise with the
level two flight time.
Other Work
The legs are finally all done for the first modular
vehicle. Getting that all right was
rather more work than expected, but everything is in CNC code now, so building
more won’t be too bad. We proof-tested
the entire module to 600 psi again now that all
welding and fabrication is done, and we should be ready for a first flight as
soon as we can take two shop days away from Quad vehicle work.
We have had all the gear and an engine on it, but still not
captured in a picture yet.
http://media.armadilloaerospace.com/2007_06_03/module.jpg
http://media.armadilloaerospace.com/2007_06_03/moduleScale.jpg
http://media.armadilloaerospace.com/2007_06_03/legBottom.jpg
http://media.armadilloaerospace.com/2007_06_03/legTop.jpg
I received our new three-antenna Septentrio
attitude sensing GPS system. This is
about a $10k system, but the capabilities were previously only available in
$50k systems, so hooray for progress! My
first impression was very good – the hardware, cable set, and serial interface
were all excellent, and it was very easy to get the system integrated with our
backup flight computer. However, there
have been issues with getting a solid attitude lock. Russ found the first problem that was causing
me grief – I had been hoping to fly the vehicle with the new system as a
passive ride along first, but to my surprise, simply plugging the antenna into
the old GPS a few feet from the new antennas caused it to lose all the
satellites. With the antenna off, I can
get good position / velocity data, but the attitude lock seems fragile, and
sometimes it never comes back to fixed integer ambiguity mode after losing
it. I am going to completely remove the
old GPS board for the next set of tests.
I really want to get this working, because the orientation
is one of the things that can be manually screwed up. The vehicle needs to be “pointed north”
before launch so it knows which axis should move the vehicle in a particular
direction. If we ever put the vehicle
down backwards, it would flip itself over immediately after liftoff, and I
really want the ability to reset the system even if it has tumbled at high
speed and hit the rate limits on the gyros.
A full ten+ minute space flight will also cause a couple-degree drift in
our current gyros, which I would like to avoid.
News
