January 15 and 19, 2002 Meeting Notes

 

In attendance:

 

John Carmack

Phil Eaton

Russ Blink

Bob Norwwod

Neil Milburn

Joseph LaGrave

 

 

Engine Development

 

The next screen pack test engine is almost ready to fire:

 

media.armadilloaerospace.com/2002_01_19/portedEngine.jpg

 

This has a 0.25” throat and a 1.1” diameter by 2” long catalyst pack, which we will be filling with alternated stainless and pure silver screens. I also finally got around to measuring our other catalyst packs for documentation: Juan Lozano’s motor has a 2” diameter by 1.9” deep catalyst pack for a 0.56” throat, but it was designed to be expandable to a 0.73” throat with the same pack. Our foam attitude motors have a 0.5” throat and a 1.65” diameter by 1.2” deep catalyst pack. We haven’t fired Juan’s motor very much, but when we took it apart there was a fair amount of silver plating dust that was coming off. His plated screens definitely react much quicker than the pure silver screens, but we are still concerned about lifetime and batch to batch variability with plated screens. If we can make a pure silver screen based pack work, it will probably be a simpler (but heavier) configuration.

 

The new test engine has a heavyweight aft section, which includes an NPT port, so we will finally start collecting chamber pressure to compare some different pack options. After we have done that, we will probably do some very brief biprop firings in a heat-soak mode. I am curious to see if we can make anything auto-ignite with 70% concentration peroxide, because that would drop the flame temperature enough that we could use an uncooled plated TZM chamber. The Isp would only be around 200, but it could be a very cheap combination with good operability.

 

The new engine also has a 0.010” relief cut on the bottom to help retain the 0.025” thick copper gasket. We will remember to re-torque the bolts after the first hot fire. We still have threads in the brass for this nozzle section, but if we use it with a flight nozzle, it will have a flange with stainless nuts

 

We are getting rather frustrated with our foam pack engines. All of them were loose enough to rattle today, even though they had only been on two flights since they were repacked. We added a few more discs to the attitude engines, and they worked fine, but our first short peroxide test of the full up lander had the main engine never clearing up. We opened it up and saw such appalling stripping in the center of the foam discs that we wondered if our peroxide was over 90% concentration. We pulled out the hydrometer that we bought a year ago (and never even tried out), but it showed right at 1.4 g/cc as it should. We think that there must have been one poorly plated sheet of foam. We added some more discs and got it to run ok. We need to stop messing with our engines to make real progress on the vehicles.

 

We will probably see if we can make a unified screen catalyst pack for the rotor tip engines and the attitude engines on the lander. We are going to see if we can live with the sizes of pre-cut stainless discs available from McMaster for the smaller engines, because the stainless is such a pain to cut. For larger engines, we may just cut them by hand with offset grip shears, or have them water jet cut.

 

I am going to try moving the injector pressure drop from the spreading plate to a separate jet before the engine to allow it to be changed without disassembling the whole engine.

 

We have emptied out our third drum of peroxide.

 

Lander Work

 

We did finally get the lander ready to lift off again after the false starts with the engines, but we didn’t have a good day.

 

The only new things for this flight were the use of the 1000mW Esteem wireless units on both sides, and the use of an anti-static wrist strap when using the joystick (there is a belief that static may have caused the joystick reset on the last crash).

 

The vehicle started tipping over after liftoff, and on coming back down it kicked off two of the foam pads and bent one of the legs. We noticed that one of the engines wasn’t canted properly, and we had seen this twist-and-tip behavior before when the cants were wrong. I later configured the simulator with that setup and was able to replicate the behavior. The flight control is supposed to pay much more attention to the non-roll axis, but when the roll gets far enough out of bounds it still starts trying to correct it. I am going to change that so if the other axis are more than 10 degrees or so off, it completely ignores roll control.

 

We have kicked the pads off on landing too many times now, so we are going to modify the legs to positively retain the foam instead of relying on adhesive. We have used adhesives of various quality (the one used today wasn’t so good), but even when we had good bonds, we have ripped the foam under the bond on bad landings. We are going to try out putting aluminum skid plates under the foam, and secure those to the top plate with sliding bolts going through the foam. A bit of skidding on a landing with sideways velocity will be better than instantly gripping.

 

We are also going to shorten the legs, except for the one in front of the pilot. The vehicle is just too heavy now for long legs in single shear. We had several successful flights of the short form lander, which was much narrower than this will be, so we think it will be a good trade.

 

We are probably going to try again on Tuesday.

 

http://media.armadilloaerospace.com/misc/jan19_02.mpg

 

 

Rotor Vehicle Design

 

We spent a while discussing our options for the tip rotor powered vehicle. The current thinking is to have fixed, steeply pitched blades on a gimbaled hub with two linear actuators for the computer to control. If it tests out ok, we are looking at rotor-on-top, ascending without aerodynamic stability, then descending with stability in the same orientation. The descent won’t be an autorotation at those blade pitches, more like a drop with thin dive brakes extended. We really don’t know how challenging the unstable control will be from a control systems standpoint. The time rates will be a lot more lenient than the pure rocket vehicles, because the rotor adds a lot of gyroscopic stability, but the required actuation forces may still be very significant for a gimbal. We have considered many options to stay stable both ways, but that involves either swapping ends at apogee and changing blade pitch, or deploying a fairly significant aerodynamic modifier for half the flight, which would have to be on the ascent phase for a rotor-on-top vehicle. Rotor-on-bottom offers some efficiency advantages, but makes a powered landing without a top over difficult. Even death-swoop maneuvers got some consideration.

 

I have ordered a big set of extruded rotor blades, and I am starting to assemble all the needed bearings and linkages for us to start testing. After we have the rotor making solid lift on the test stand, we will probably attach it to the top of the current lander. We could optionally make short flights using the attitude engines for correction until the gimbal is working.

 

 

New Software

 

I am working on a unification of several of the software programs we use. The simulator, remotePilot, graphLog, and gpsView are all going to become a single program with much better capabilities. There are currently two or three different instances of a 3D rendered view and scrolling graphs, all with different options. I will also be logging more data over the telemetry, so we can try filtering our accelerometer signals. I will probably make the change after we get the laser altimeter integrated, where we would have to change the telemetry format anyway.