News Archive

September 8, 2001 Meeting Notes

September 8, 2001 Meeting Notes

In attendance:

John Carmack

Phil Eaton

Russ Blink

Neil Milburn

Joseph LaGrave

We got a huge amount accomplished today.

Phil and I drained about the last of the peroxide out of our drum on Tuesday, but 2000 pounds of peroxide was supposed to ship out for us yesterday, so we should have plenty in a couple weeks. We will be getting 3000 pounds more some time later.

Loading ten liters into the lander took over twenty minutes last time, between drawing the vacuum and loading it through the small solenoid, and also caused the solenoid to overheat enough that it caused a peroxide burp, so we took some major corrective action today. Our 3gpm, 30psi Teflon pump is supposed to be here in a week and a half, but it was worth doing the changes just for the next couple tests.

Russ brought a big AC powered vacuum pump to replace the little battery operated ones we had been using. This draws a better vacuum, and drastically faster. A brought a bunch of 1/2" manual ball valves that we could re-plumb the fill cart manifold with. We would up just putting one on the peroxide loading side, because the nitrogen loads quick enough through the solenoid, and the vacuum pump is probably the limitation on the suction side. I also got us a bunch of 1/2" ID clear PVC tubing to load the peroxide with. These changes brought the peroxide loading speed from 37 seconds a liter (at best, for the first liter) to ten seconds a liter, a very welcome improvement.

Our new Fike burst discs and holders came in. These are really nice, and exactly what we wanted – a male pipe thread going directly to the burst disc holder, then either an integral diffuser, or another pipe thread for a vent pipe. We got two 1/4" ones for the small lander and test stand, and two 1/2" ones for the big lander, and eventually our bigger test stand. We had been using a relief valve on the big lander, and shuffling an NOS burst disc with an adapter between the other tanks. The new discs are all 1500 psi burst, which is half of what the NOS discs are set for, so that gives us an additional safety margin.

We got all the pilot gear mounted on the big vehicle: an aircraft seat belt, the joystick, and the twist throttle. We still need to do some moving around and bracing, but everything is there and functioning.

I made a special version of our graphical flight simulator that takes forwarded joystick and throttle data from the flight computer, so we were able to stand on the pilot platform and fly the simulator on the laptop with the real vehicle controls.

We got the electronics box mounted on the correct side, but it is just hanging on by its mounting tabs, so we should probably get some bracing all the way underneath it.

I have added two switches to the test stand electronics box so we can open and close the ball valve with momentary toggle switches, instead of the swap-the-alligator-clips-and-pulse-the-computer silliness we had to do before. Both ball valves now have the same CPC connector, and can be plugged either into the flight computer or the test stand box.

Russ drilled the upstream vent hole into the ball valve. We have been fine with the non-vented ball for the test stand work, because the valve just opens and stays open until the tank is blown down, but on the vehicle, the throttle will be opening and closing, so we needed to make sure we never trapped a ball full of peroxide. It wouldn’t likely be a problem sitting there for a little while, but if it did start to catalyze for any reason, like heat soak, the ball and valve would probably shatter like a grenade. All that you need to do to avoid this is have a small hole drilled from the tank inlet side into the ball center passage.

Phil finally found a supplier that would make us pure silver foam without a multi-thousand dollar setup fee: Astromet is going to make us enough foam for our four attitude engines for $512. If those work out well, we may get a larger batch for all our future needs. Our current silver plated nickel foam only seems to last for about a half dozen hops in the lander attitude engines, so we hope to get something a lot more long lasting.

Phil also got us some pricing information for big filament wound tubes from Beadle Plastic: 2’ diameter tube is about $40 per foot, plus shipping. We are getting 20’ of it, so we can experiment with two 10’ long airframes. We are also getting quotes for the composite parts for the next-gen (supersonic) manned vehicle.

Engine Tests

media.armadilloaerospace.com/2001_09_08/SmoothBigEngine.xls

We changed two things for our first engine test: Russ made a new cavitating venturi with a 0.125” throat, and Phil cut additional catalyst discs to fill the area between the inlet fitting and the spreading plate. The area was pent-roofed, so he had to cut several progressively smaller discs.

The first CV that we tried was a 0.26” throat, which we believed to be too large to actually cavitate. We wanted to have one small enough this time that it positively would be the flow restriction.

The first test run sounded smooth for the first time ever on this motor, but I screwed up and aborted the data logging before it got saved out. We did another identical run (two liters at 400 psi), and got the data properly. It was perfectly smooth, and made 175 pounds, just about exactly the jetting we want for the piloted lander central engine with a bit more pressure.

We had a smoothness that we were thrilled with, but we didn’t know if it was due to the CV or the extra catalyst, so we took the CV out and let it run unrestricted for another two liter / 400 psi run. This run was also perfectly smooth, and made 400 pounds of thrust. The slight buzz every other sample in the data is in the data collection hardware, not the engine. There is a strange issue with the RS232 voltage needs of the electronics, where if I have the laptop running on battery it works fine, but if it is plugged into an AC line, it gets a buzz in the A/D signal. I forgot to unhook it for this test.

So, the cause of all of our rough runs was having a moderately large open volume between the plumbing fitting and our spreading plate. Problem solved. In the future, we will probably want to design the motors to absolutely minimize that space, because adding catalyst up there is probably limiting our maximum thrust by requiring some of the flow past the spreading plate to be gaseous instead of all liquid, making that likely the major flow limiter of the engine instead of the nozzle.

It is interesting to compare the thrust falloff vs blowdown pressure between the two runs -- the unrestricted one falls off pretty much directly with pressure, while the restricted one falls with the square root of the pressure.

Flight Test

The big vehicle is set up exactly as if it was piloted now. We ballasted it with a heavy punching bag (donated by Christian Antkow) and a water jug so that it has the same weight distribution as it will with a person on board. The total weight is now over 300 pounds dry, with a significant chunk of it well off the centerline, so it would be a strong test of the new attitude control software. The water jug will add a bunch of sloshing to make things even more interesting.

The first test was with six liters of peroxide, pressurized to 500 psi. This was our first test with the large engine and ball valve, so I rolled the throttle on pretty slow. It only got a few inches off the ground before it used up all the peroxide.

For the second test, we loaded ten liters of peroxide and increased the pressure to 600 psi.

It worked great! I kept it very low to the ground, but the flight control software easily dealt with the offset CG, and I could have put it anywhere I wanted. I am still sort of amazed that it seems to be working as well in real life as it did in my simulations.

http://media.armadilloaerospace.com/misc/Everlast.mpg

The pilot’s joystick is flopping all over the place while it was flying…

Ten liters of peroxide gives seven seconds of lift in this configuration. We can load twenty liters with the current tank, and we can get much longer tanks of the same diameter if we want to do any long duration flights.