November 13 and November 17, 2001 Meeting Notes

 

In attendance:

 

John Carmack

Phil Eaton

Russ Blink

Bob Norwood

Neil Milburn

 

On Tuesday, we were testing a couple minor modifications to the flight control software that should make it a little easier to manage the throttle.  I reduced the turnaround delay for the motor drive from 100 msec to 50 msec, and implemented a scaled remapping of the throttle to ball valve position.

 

We did a quick hop for checkout, then loaded up for a full length run.  Shortly after liftoff, a plume of peroxide started escaping from one of the engines.  I set it down, and in the process of venting of the rest of the peroxide, one of the other engines started doing it.

 

We knew pretty quickly what happened – the four attitude engines aren’t completely interchangeable with the tops, and two of them were misaligned enough that we couldn’t get one of the bolts in it.  The O-ring had blown out at the point of the missing bolt.

 

We drilled the bolt holes in the tops out slightly oversize, which allows them to fit in any orientation, and we also replaced the rubber O-rings, which got very hard after firing, with high temperature silcone O-rings.

 

Since we were taking the engines apart, this was also a perfect time to test the chem.-etched injector spreading plates that ERPS had made up for us.  The micro-perf sheets that we had been clamping at the tops of the motors was 7% open area, which was quite a bit too much to provide enough pressure drop for smooth engine operation.  The ERPS plates are thicker, and have far fewer holes in them.

 

We couldn’t get test stand data because of a problem with my new laptop, but we ran it once for a basic check, and it seemed to run well.

 

By Saturday, I had addressed the test stand issue, but I’m not real happy with it.  Microsoft changed the behavior of the serial DTR line with windows ME, where previously it was off when not in use, but now it stays on from the time you boot.  Our test stand electronics use the DTR line to trigger the test valve, so as soon as it was hooked up, the valve would stay open.  In looking around online, several other devices, like radio push-to-talk triggers, are also having problems with this.  Microsoft apparently has a fix for it, but you need to sign up for a support contract to get it.  For whatever reason, the couple fix programs I downloaded online only temporarily cleared DTR for me, so I had to try and work something out myself.  I eventually found an awful “rain dance” that does solve it – run a little program that clears DTR and spins in place, run the test stand program, which will fail with a sharing violation trying to get the serial port, the ctrl-C break out of the stub program, and DTR will stay off for the rest of the session.  Blech.

 

We had a couple issues getting the runs working properly on the test stand, but when we finally got it all worked out, the small engine spreading plate turned out to be undersized enough to cut the output from 60 pounds to 30 pounds of thrust, and the big engine had a bit too much injector area, because the thrust still tracked the pressure blowdown directly, instead of falling off with the square root of pressure, as usually happens with sufficient choking.  However, both engines are operating smoothly, so this seems to be a pretty good combination.  We were having some solenoid trouble on the test stand, so the small engine thrust may not be correct.  In any case, we don’t mind losing thrust on the attitude engines, because we still have an excess of control authority.

 

The Doppler radar speed sensor that we had been waiting for showed up on Friday, and I was all set to push really hard to try and get auto-hover and auto-land operational for Saturday, but it turns out that the sensor only returns the absolute value of speed, not including the direction.  That makes it useless for our purposes, but I am hoping the company (GMH Engineering) can modify one for us.

 

We did a couple test flights, and everything worked fine with the control changes, but it is still harder to smoothly fly and land with the throttled central engine than when we were using the attitude engines for throttling.  I am hoping that computer control will work all that out when we get a speed sensor.

 

We did have some link quality problems with the 802.11b link between the laptop and the flight computer, so I moved a little closer outside to get it to clear up.  I have ordered two full-watt units from Esteem , which should give us several miles of range, even with omni antennas.

 

Bob brought the new dual tank frame over so we could adjusted the seat position for proper balance.  Our new tanks should be here soon, so we will probably be flying this vehicle relatively soon.

 

Mark Henry, the ex-propulsion lead for Beal Aerospace stopped by on Saturday to talk with us, and we got a lot of good tips and information from him.  A few notes:

 

He favors a special silver plating process on stainless screens, including the samarium nitrate coating that some literature had reported as unnecessary.  He said they could get effectively unlimited catalyst life, and 40 msec startup times.  Apparently the difference between pure silver screens, which we found to not work very well, and the special plating process is quite profound.  We are going to buy some packs from him to experiment with.

 

Anti-channel rings should be fairly thick, and made out of the same material as the motor casing so that they expand with the case, rather than the pack.

 

Silica-phenolic ablative nozzles were heavy, but very easy to make and work with.  Carbon-phenolic nozzles lasted less than 20 seconds with peroxide engines.

 

He was fairly down on hypergolic miscible fuel peroxide combinations, mentioning the sever motors that have been blown up in experimenting with that.

 

If you are setting up a really large-scale concentrator, 90% peroxide comes out to only about $0.50 a pound.