May 26, 2001 Meeting Notes

 

In attendance:

 

John Carmack

Phil Eaton

Russ Blink

Bob Norwood

 

 

New equipment:

 

Hose reel

Face shields

Two two-liter flasks

12VDC to 115VAC inverter for remote test stand

-10 AN Teflon hose, hose ends, and fittings for 600lb engine test stand

Four NOS Pro-Race solenoids for manned attitude engines

 

We did three engine tests today, and then ran two loads of peroxide in the VTVL.

 

media.armadilloaerospace.com/2001_05_26/may26_01.xls

 

The first engine test was a proper 3x expansion nozzle for a half inch throat. Having made almost 40 pounds with a nozzleless engine earlier, we were expecting to see almost 50 pounds of thrust. The run was very ragged, and didn’t make much more power at all. We ran it again, with similar results.

 

We are assuming that the catalyst pack was messed up when it was transferred from the old test stand engine (which now has the pressure port plug cross-threaded) to one of the engines we pulled off the lander for this test.

 

The third test was of a quickly machined plug nozzle (aerospike). This was sort of a throw-away test, because we know the throat areas aren’t comparable, so we wouldn’t be able to say anything conclusively about it. It made less thrust, and was still rough. We have designs for better plug nozzles that we will be testing later.

 

We need to do more investigation of this poor performance today. I suspect channeling around the pack, which probably wasn’t very true after being pulled out of one engine and pushed into another.

 

Based on suspicion of moving catalyst packs in and out of the engines, we made all new catalyst packs for all the VTVL engines instead of trying to stuff the original one back into the test stand engine when we moved it back to the VTVL.

 

Russ took a quick stab at making the microcontroller modifications to cut some latency out of the gyros, but we didn’t have enough time to work them out, so we went back to the original code for the sensor board.

 

We tried the first load with the same flight control as was used on May 15^th. It was a little windy, but not as bad as the last time. I hopped it up several times, but it couldn’t stabilize itself. Using our new graphing software, it was easy to see that it just didn’t have the control authority.

 

The next load, I changed the flight control parameters so that the minimum delay between attitude corrections would be lowered to two instead of three. I expected to see more overcompensation (although still less than on May 12^th, because of the addition of rate extrapolation), but it would give it about 40% more control authority.

 

I lifted it off and flew it for the entire propellant load of seven seconds, and it stayed stable even when it drifted sideways enough to pull one of the tethers taught. Yeah!

 

media.armadilloaerospace.com/2001_05_26/sevenSeconds.mpg

 

The control graph does show reasonably large oscillations, but it kept it under control.

 

media.armadilloaerospace.com/2001_05_26/sevenSeconds.gif

 

We probably won’t be flying the VTVL again until we are converted over to the fiber optic gyros, unless the delivery time is several weeks off, in which case we might go ahead and work on roll control with the existing gyros.

 

After we get the FOGs and roll control in and tested, we will be taking off the tethers and flying it around free.

 

Manned Vehicle Notes

Bob is looking into the local supplier of pure carbon fiber NGV tanks, but if nothing better turns up, we will probably use Neil’s tank.

 

We need to take some peroxide to Bob’s shop so we can test its reactivity on various metals and coatings to be considered for the vehicle.

 

The primary design options are:

 

One tank in center, pilot offsets CG. Simplest and strongest, but requires the most attitude control.

 

One tank slightly offset from center, balanced CG at blowdown, when attitude control has least authority.

 

Two tanks, pilot in center, CG stays balanced throughout flight, but with a larger polar moment.

Even though we are intending to start with a standing pilot, this would be the direct path to a supine pilot / high G vehicle.

With two of the tanks that Neil found, the vehicle would be too heavy to lift off unmanned with just the attitude jets, so testing couldn’t be quite as incremental. If we get carbon fiber tanks, it might still be ok.

It would require a significant additional amount of plumbing to join the tanks at the top and bottom.

 

I think we are settled on pilot-facing-between-cross members, rather than pilot-looking-down-cross member, because it lets all the cross members be fully triangulated up the tank and out to the attitude engines. This joystick orientation will change the control logic a bit, but I don’t expect it to be a problem.

 

The driving design issue is the attitude control system. The NOS solenoids are still the best valves we have found in our pressure / response / materials domain, and that limits us to about 50 pounds thrust at 400 psi tank pressure with the 8.5 amp solenoids, or 70 pounds with the 30 amp solenoids. If we need more thrust than that, we would need to either gang multiple solenoids, add more attitude engines, raise our tank pressure (limited possible gain), or move to slow acting servo valves for attitude control.

 

Ideally, the attitude engines would be mounted vertically at the height of the CG, which would apply all of their thrust as a rotational torque. If we decide to mount them completely below the pilot, we could regain the torque be canting them inward, but they would be less efficient at providing lift, which would be a bit of an issue if we used pulsing-off instead of pulsing-on for attitude adjustments.

 

If we have a 200 pound mass six inches from the centerline, the attitude engines would need to be around eight feet from the centerline to be able to keep the vehicle from rotating at the lowest engine blow down pressures, and that wouldn’t give much authority for wind. We can help by placing the electronics box (10 pounds) out by the opposite engines, and that could be extended to actually adding some ballast at the far end. It would probably be reasonable to completely statically balance the pilot with less than twenty pounds of ballast added to the electronics box.

 

We need to be conscious of transportation logistics. A very wide vehicle will probably require the outer legs to be removed for transportation. What is the maximum dimensions we can fit in our existing vehicles (Phil’s Suburban or Neil’s van)? How wide could we put something on a custom trailer?

 

Landing / floatation: foam blocks are working out great on the demonstrator, but the manned vehicle will be about ten times larger, so it bears reexamining. A large inner tube is still a contender.

 

Pilot orientation: leaning forward against the tank, or leaning back against the tank? Leaning back against the tank almost makes the entire vehicle a big “rocket pack”.

 

 

To Do List

 

Pick up more peroxide from Rinchem. Phil, let me know if you want me to come over to help, or if you and Russ will take care of it.

 

Get some more silver plated foam. Phil mentioned getting it done at a level in between the two ones we have tried, but if it hasn’t already been started, I would probably prefer to stay with the last version we got, and just change other pack variables.

 

Finish development on 50 pound thrust engines. Russ, you might want to make the converging angle go deeper so the throat isn’t as long, but it probably doesn’t matter much. We might want to investigate other engine issues, like external threading to make pack insertion easier, designing for O-rings on both closures, etc. before making all four nozzles.

 

Investigate pure silver foam for catalyst packs. We may be able to get by with a single layer of thick silver foam instead of stacks of discs, which would reduce variability.

 

Investigate using external threads on the motor tubes to make catalyst insertion easier.

 

Investigate using an injector plate with lots of small injector holes, both with and without inert foam spreaders. I would prefer to get this done in an automated way, rather than doing it by hand in a drill press. We should figure out exactly what plate diameter and hole diameter we want, and see about getting Darin’s friend at the laser jet shop to cur it.

 

Test 30+ second burn on engines.

 

Test pack wear conditions. The only way we are going to be able to do this is if we make a completely separate engine for it, that will never get cannibalized for anything else. We should fire it a few times a week, and just see how it degrades over a couple months.

 

Aerospike testing. Technically, they are plug nozzles, not aerospikes. We need to do an exact, back-to-back test with the same throat area and a chamber pressure sensor.

 

Fiber Optic Gyro integration when they arrive.

 

Start development on 600 pound thrust engine.

 

Vertical test stand. Something good for several thousand pounds of thrust. Neil: you mentioned before that you knew some people handy with concrete and construction that we could probably get cheap. Still an option?

 

Closed loop servo valve control.

 

New electronics box.

 

Basic GPS + accelerometer integration for landing assist.

 

Move the test stand hardware into one of the small waterproof electronics boxes, and see if it is possible to run the engine with a spare serial control line, instead of requiring both a serial cable and a parallel cable.

 

Roll control on the lander.

 

Parachute canisters and fins for the ballistic flights of the lander with the larger engines. I am thinking something like a PML piston ejection system. I would prefer to investigate cold gas ejection, but we could do some flights with blackpowder ejection charges.

 

Drill ball valve vents. Practice on our galled valve, then drill the servo valve. We need to decide for sure which side is inlet and which is outlet before doing this.