June 5, 2006 notes
The tall vehicle (internally called the VDR for Vertical
Drag Racer) is just sitting here, completely ready to fly, but we havent
secured a new test site yet after annoying a neighbor at my 100 acre site last
month. Negotiations at McGregor seem to
have hit some unknown snag, so we are now looking at going back to the Oklahoma spaceport for
the tethered tests. They are a definite
no-go for any vertical flight operations, but we should be able to do our
tethered tests there, and we already had a business relationship with them, so
hopefully we can get something going fast.
We dod do a few more
modifications to the vehicle this month:
We sprayed fastblock-800 (http://www.kirkhill-ta.com/pdf/FB800-Fastblock_800_Series.pdf
) on the bottom of the lox tank, because it is fireproof. We used cheap spray polyurethane foam to
insulate the top half inside the intertank section,
where we werent worried about the insulation catching fire. The fastblock was
several years old and past its use-by date, but it still worked fine. The bottom layer isnt very thick, but it
should still help. After we set our
electronics box, insulated by a conventional foam, on
fire on the test stand once, I am a little paranoid about flammable stuff on
the vehicle. The Flowmetrics
guys reported that they use cheap fiberglass insulation with a glass/epoxy top
coat on their lox tanks and plumbing, which sounds like a good idea for the
future, but all the legs on the current vehicle made spray application
convenient. We did have a bit of an
issue with the polyurethane spray foam on the inside, where some of it didnt
get mixed well in the nozzle, allowing some uncured gunk to leak out the bottom
of the intertank onto the outside foam.
We added ladder rungs to the legs by the hatches so we can
climb up and turn the computer on and fill the high pressure bottles without a
stepladder. This is the type of little
thing that causes weight growth, but we consider the operational convenience
We upsized the propellant valves from ½ full port to ¾
reduced port, which should be a 2x flow increase at a given pressure drop.
Our next generation electronics board is all laid out. It will probably take somewhat over a month
before we have one together well enough to test with, but it will be giving us
improved capabilities and it should offer increased reliability.
The VDR can easily make the 90 second flights for the level
1 lunar lander challenge, and it might make the level
2 flights if we get our engine Isp up to a reasonable
level. However, it has the operational
disadvantage of needing a lift truck to erect it and move it around, and the
high length to base ratio doesnt lend itself well to landing on rough terrain. We are still pretty confident that with
ground contact sensors shutting the engine off immediately that we wont have
tipping conditions like at the X-Prize Cup, but it is still certainly something
that can go wrong. The 65 diameter
vehicle would easily handle level 2, but it is a really big vehicle, and
testing will be a lot more trouble.
We decided to build a completely new vehicle specifically
for the lunar lander challenge. It consists of four spherical tanks with a
gimbaled engine in the center, giving a very, very low CG (below the gimbal point, so I am going to have to flip the actuator
calculation), and the top crossmember is conveniently
located to allow lifting bars to be stuck in so a few people can move the
vehicle around by hand, which is a pretty significant operational benefit.
Matt made some nice renderings of the vehicles
recently. The technical details arent
perfect, but they are pretty representative. The caption on the 65 cutaway is
We need a vdr_flight to complete
Total vehicle weight on the quad will be a bit lighter than the
VDR, probably around 600 pounds, and the propellant load will be 2100 pounds at
66% full, giving a mass ratio of 4.5.
That only needs a delivered Isp
of 133s to hover for 200 seconds, so we have tons of margin.
One of the key determinations for this design is that the
last hold down test showed that a fairly severe pressure blowdown
and mixture ratio change didnt hurt the engine at all, so we are really happy
to ditch all the high pressure bottles and regulators. This makes ground operations a lot simpler,
and removes several possible safety issues.
Isp will probably
drop pretty far at the low pressures towards the end of the burn, but we can
make up for that with mass ratio.
This vehicle will have 4x the drag of the VDR, so it wont
be able to fly very fast, but it might still have some other uses. We will probably put a seat on top, Im sure
someone will want to go for a ride
Interestingly, it should also float, and could probably liftoff just
fine while bobbing in a lake.
The bottom of each sphere has a combination sump / shock
mount welded on. After I made the first
prototype, I realized that a hydraulic shock screwed into a lox tank would very
rapidly become hard as a rock, so I redesigned to allow a phenolic
spacer to be screwed in between the tank mount and the shock absorber. This was my first project really setting up
for a mini production run on the new mill, and once I got it all debugged, we
made eight of them in pretty short order, so we have a complete spare set. This was also my first time using a thread
mill, which I like very much. I am
considering threading a lot of large diameter things now, especially the engine
chambers onto the injectors.
Here are all the sensors and actuators external to the
electronics box, laid out for connector wiring.
Lox valve, fuel valve, range safety valve, four roll thrusters
solenoids, two engine purge solenoids, two engine igniter solenoids, two gimbal linear actuators, fuel pressure transducer, lox
pressure transducer, and chamber pressure transducer. Note the sealed heat-shrink boots on the
Since we are going to be doing a lot of tank welding in the
coming year, we upgraded to a brand new, very high end welder, a Miller dynasty
). We had a couple pretty good size
welders, but they couldnt run continuous duty at the amperage required to weld
¼ thick aluminum hemispheres for 20 minutes at a time, let alone the big 3/8
thick hemispheres (and I now wish I had gotten the big ones in ½ thick). James and Russ really like the new welder,
and it is actually a lot smaller and lighter than the much less capable welder
it is replacing. We had previously
upgraded to a water cooled torch and an automatic wire feeder in preparation
for this, and using the old mill fourth axis as a rotisserie has worked out great,
so we are pretty much set for fabricating full size production vehicles now.
We are using stitch bands to pull the hemispheres together
now for the initial tack welds. It sure
beats trying to get them closer together with a hammer.
James and Russ spent a long time trying to automate the tank
welding process with a jig for the torch and wire feed, and an idler wheel to
keep everything lined up exactly in the weld groove. Unfortunately, they werent able to get it to
work perfectly. When it was going right,
the weld looked amazingly good, like a friction stir weld. However, the wire feed just didnt seem to be
consistent enough to do the job. It was
improved by cleaning the feed tube up and supporting it in a very gentle arc, but
it still didnt quite cut it. James is
probably going to wind up holding the torch and wire feed by hand while the
tank rotates so he can make all those minute adjustments that good welders do
while they are working. This should
still be a good improvement over the previous way we did it with manual
rotation of the sphere and individual welding rods.
When I was making a new injector head for the quad engine, I
discovered that the end mill that I was using to cut the injector holes was
going a little bit too far down, leaving a countersink at the exit. We know this really hurts performance, so
that explains why the Isp
wasnt good on the last engine.
Unfortunately, it may also be why the last engine didnt melt the
injector head, so we will need to be careful testing the new engine. The new engine has slightly more injector
area, a slightly leaner mixture ratio, and slightly less film cooling.
The vehicle is almost done, and we are probably going to
build a second, nearly identical one and run two quads at the XPC for the level
one and level two prizes. For level one,
we wont need to refuel at the midpoint.
We will probably take the VDR to the show, but not fly it. We discussed possibly using it to compete in
the suborbital payload contest, but unlike all the other parachute recovered
rockets, we would need an experimental permit due to the long burn time, and we
dont have a valid test site (or the time) to get that wrung out before the
We had the pleasure of having Lutz Kayser,
the principle behind the OTRAG project, visit our shop this month, and he
brought along some of the actual thirty year old hardware from the program for
us to look at. I have been corresponding
with Lutz for a few months now, and I have learned quite a few things. I seriously considered an OTRAG style
massive-cluster-of-cheap-modules orbital design back when we had 98% peroxide
(assumed to be a biprop with kerosene), and I have
always considered it one of the viable routes to significant reduction in
orbital launch costs. After really going
over the trades and details with Lutz, I am quite convinced that this is the
lowest development cost route to significant orbital capability. Eventually, reusable stages will take over,
but I actually think that we can make it all the way to orbit on our current
budget by following this path. The
individual modules are less complicated than our current vehicles, and I am
becoming more and more fond of high production methods
over hand crafter prototypes.
There were a lot of clever design aspects in the details of
the parts he showed us. Im not sure how
widely he intended the details to be spread, so I wont go over them, but I definitely
got several things worth thinking about from them.
The most surprising thing is that when he left, he gave us
one of the injector assemblies to try and convert to our propellant
combination. I have been really amazed
at his generosity. OTRAG used maximum
density acid / kerosene, which has a much, much higher O:F
ratio, so we will probably have to plug 2/3 of the oxidizer holes to run with
lox, which wont be optimal, but it will still be interesting. The vehicle work for the lunar lander challenge is our top priority, so Im not sure when
I am going to be able to fabricate the manifolds for the injector, but it is
definitely on the list of things to do.
Our shop is getting a bit crowded:
I mentioned this once when we first moved there, but we do
have an official Armadillo forum at http://www.spacefellowship.com/
where public questions can be put to us.
I try to answer questions in email, but I prefer to make answers part of
the public record to reduce repeat questions.