May 6, 2006 notes

 

Horizontal Testing

 

We continued to have problems with the current engine melting lots of wires on the vertical test stand, so we finally broke down and made a horizontal test stand. I have resisted this for a long time for two reasons: When possible, you should test the same way you fly, and we fly with the rockets pointing down. More importantly, it is easier to have catastrophic hard starts in a horizontal configuration, because a significant amount of propellant can pool inside before running out the nozzle. Several years ago when we were testing 90% peroxide / kerosene biprops, we had a hard start due to a leaking solenoid letting kerosene seep into the catalyst pack. After collecting all the pieces from that, I swore off horizontal test stands. At this point, our torch ignition system and pressure transducer interlocks seem to be reliable, so we are guardedly going to do some development work horizontally.

 

When we fired the engine horizontally, it was clear why we were having heat problems: the plume was 35 feet long! The film cooling on the current injector (20% of fuel) was, if anything, far more effective than necessary, since the graphite throat really didn't get very hot after a couple twelve second burns. The film cooling shows distinctly in the exhaust plume as the cloudy layer around the mach diamonds.

 

http://media.armadilloaerospace.com/2006_04_08/horizontal1.jpg

http://media.armadilloaerospace.com/2006_04_08/horizontal2.jpg

http://media.armadilloaerospace.com/2006_04_08/horizontal3.jpg

 

http://media.armadilloaerospace.com/2006_04_08/horizontalFiring.mpg

 

We are going to angle the engine down into the water more in the future, and add some water spray directly into the nozzle to reduce noise.

 

I had been closing the valves with a lox lead, so the fuel always shut off last. This was intended to prevent raw lox from ever shooting at the hot graphite nozzle, but the effect was to always give us a big puff of flame on shutoff, often leaving some burning fuel backwashed onto the motor. I am now shutting them off simultaneously, which avoids the flaming fuel at the end of a run.

 

Our carbon-fiber-with-special-binder reinforced graphite chamber / nozzle combination from Cessaroni Aerospace is in, but we haven’t fired it yet:

 

http://media.armadilloaerospace.com/2006_04_18/reinforcedGraphite.jpg

 

With film cooling, I have little doubt that this will run for 360 seconds, but the latest rules for the Lunar Lander Challenge will allow us to swap chambers between the two 180 second hops if necessary.

 

Hold Down Test

 

We set out to our 100 acre test site to do first a hold down test, then a hover test under the lift. We had four concrete blocks cast with a cubic yard of concrete each (4000 pounds) and a lot of internal reinforcement to act as the hold down points.

 

We had an initial problem with getting the engine started. It turned out that the last test we did with pressure loading water into the fuel tank left a good bit of water in the fuel line to the engine. We had blown the entire helium load out through the engine, but the feed line is a corrugated 1” line, and the gas velocity just wasn’t high enough to entrain all the water. We had the vehicle up in the air, but the igniter just wouldn’t start. The safety interlocks worked fine, so it never did anything other than make a tiny puff of vapor come out the bottom of the engine, but even after a dozen attempts, it wouldn’t light. We let most of the pressure out of the tanks and investigated, finding that the fuel coming out of the solenoid barely smelled of alcohol, so we purged the lines until it seemed that we had gotten all the water out.

 

We re-pressurized and lifted the vehicle up again, and the engine lit right off. I throttled up, and it immediately throttled back down to idle. We had left a little bit of slack in the chain to the blocks, with the intention that it would lift up a few inches or so and come taut on the chains. Unfortunately as it pulled on the first chain, it rotated the vehicle enough to trip a roll abort condition, which put it into auto-land mode. This has been happening more and more to us – there are so many automatic no-go, abort, and shutdown conditions that most of the time we try to test something, I wind up having to disable something. This is probably a good thing.

 

Joseph went back out and carefully lifted the vehicle up until the chains were taut, and we started it again and ran it until lox depletion. The fact that it was lox depletion was a bit of a surprise – we expected it to run out of fuel first. When we analyzed the telemetry, it was clear what happened. The roll thrusters were going essentially 100% duty cycle, so the tank pressure dropped a fair amount, but the lox had been sitting there long enough while we debugged the ignition issue that it was self pressurizing to a fair degree. This told us a few interesting things: the burn was still stable with the fuel pressure less than 10% over the chamber pressure, and the film cooling kept the ablative chamber and graphite nozzle from eroding even though the mixture ratio went from fairly rich, through stoichemetric, to very lean at the end of the burn.

 

http://media.armadilloaerospace.com/2006_04_18/holdDownTest.mpg

 

Unfortunately it was too late, and we didn’t have the helium for pressurizing again, so we weren’t able to put the vehicle on the stands for a hover test. All the telemetry looked perfect, so it very likely would have flown fine. I’m sure we’ll have to tweak the gains a bit, but this vehicle is probably capable of doing the 90 second flight legs for the level one lunar lander challenge right now.

 

The second unfortunately was that we pissed off a neighbor with this test. Our site is 100 acres, but based on the decibels at the point we were at, it was probably still 102 db at the property line, which I agree is probably excessive for unsympathetic neighbors. Doing it at 8:45 in the evening probably didn’t help, either (we didn’t expect to be that late…). Testing rockets is a dubious enough activity that any determined person could probably get us shut down on some grounds, so we aren’t going to push them on the issue. We are looking at leasing some land down in McGregor, TX, right next to the Space-X test site. If their big engine tests don’t bother folks, they won’t even notice us. J

 

We have 4000 pounds of anchor chain to use as the "tether" on our upcoming flights without the lift. We will let the vehicle fly with a reasonable length of 6000 pounds strength steel cable connected between a leg and the anchor chain, so if the vehicle takes off uncontrolled for any reason, it will just jerk up a few chain links, then progressively lift a heavier and heavier length of chain until it rotates over and comes crashing down. The vehicle will be destroyed, but the damage will be strictly localized.

 

http://media.armadilloaerospace.com/2006_04_18/anchorChain.jpg

 

 

Vehicle Work

 

The 65" tanks are in. They are 65” ID, 3/8” original thickness 5086 aluminum, and weigh just under 400 pounds per sphere. The 36” ID spheres we have been using are 81 pounds dry, 950 pounds full of water, and burst at 760 psi. By direct scaling, the 65” ID spheres should hold 5.86 x more volume, and burst at 640 psi if the weights are distributed equally.

 

http://media.armadilloaerospace.com/2006_04_18/65InchSphere.jpg

(note that if you cut the legs down a bit, you cut fit the entire X-Prize Cup vehicle in the background inside the sphere...)

 

We have made a big band clamp to sinch the hemispheres together, with holes for tack welding every two feet or so. We were able to bang the 36” spheres pretty close by hand, but the 65” spheres didn’t have as close of tolerance.

 

I realized something useful -- the old mill has a very stout rotary chuck on it that we are going to be able to adapt to rotating tanks for welding. All we have to do is turn it sideways, crank the table down, and rig up some sort of giant tailstock for it, and it should be able to spin even the big tanks. We bought a wire feed gun and a water cooled torch, and we have a new high end welder coming in soon, so we expect to be able to weld complete tanks with one continuous, high quality bead.

 

Matt made a rendering of what the final 65" tank vehicle will look like:

http://media.armadilloaerospace.com/2006_04_18/challengeVehicle.jpg

 

The differentially throttled 65” diameter vehicle is our current bet for commercial service, but it looks like we are going to be detouring and building a different vehicle first that is specifically targeted for the Lunar Lander Challenge. We might actually have three flying vehicles at the ‘Cup this year, but that is a lot of work still ahead of us…

 

Random wiring note: I have started using the shrink sleeve strain relief for CPC connectors instead of the screw-down clamps. I like them, because they provide two significant advantages: Used in conjunction with the CPC receptacles with an o-ring in the bottom, you get water-tight wiring, and the strain relief doesn’t unscrew. It has happened too many times to be coincidence: every time someone new starts working with CPC connectors, they will inevitably at some point unscrew the strain relief instead of the plug, twisting all the wires badly.

 

A couple other random pictures:

http://media.armadilloaerospace.com/2006_04_08/makingChips.jpg

http://media.armadilloaerospace.com/2006_04_18/nomexSuits.jpg

 

Warning: I am swamped the next month, and any update may be late.