January 29 and February 2, 2002 Meeting Notes

 

In attendance:

 

John Carmack

Phil Eaton

Russ Blink

Joseph LaGrave (Saturday)

 

 

We did a lot of engine test firings this week with the 1” diameter by 2” long catalyst pack motor.

 

Tuesday tests:

 

media.armadilloaerospace.com/2002_01_29/combined.xls

 

1: Initial pack length of 1.5” (two 0.25” spacers), 6/64" jet, low pressure.  Cloudy result, but smooth thrust.

2: Reduce jet to 5/64", same result, although clearing a little towards the end.

3: Replace one 0.25" spacer with more screens.  Clean run.

4: Double pressure.  Thrust doubles.

5: Go back to 6/64" jet.  Little increase in thrust.

6: Increase jet to 8/64".  Little increase in thrust.

7: Double pressure again.  Thrust doubles again.

8: Long run at high pressure.  Stays level and smooth the entire time.

 

All screens were 20 mesh, alternating stainless steel and pure silver. We bought pre-cut stainless screens from McMaster, which makes life much easier. We also found that using a soft plastic underneath the silver screens when we are punching them works much better than the hard polycarbonate we were using before.

 

We were not using a spreading plate at the top of the pack at all, just relying on the screens to spread the peroxide and the jet in the fitting leading to the engine to meter the flow. This may not scale to larger motors, but it worked very well on this one.

 

These were probably the smoothest runs we have ever seen, but I am rather mystified with some of the results.  Increasing the pressure seems to give a directly proportional increase in thrust, but the thrust stays constant across the entire run, even on the last run where the ullage pressure decayed nearly 100psi.  The thrust didn't even vary with the square root of ullage pressure, as we have seen before on tightly restricted motors.

 

The only thing I can think of is some interaction with the solenoid, so we will do future tests with the big ball valves.

 

While it was curious that the first clear run produced exactly the same thrust and smoothness as the two previous cloudy runs, it did run for half again as long on the same amount of peroxide.  That is probably a pretty good failure mode for an engine -- when it clouds up, your Isp drops from about 115 to 75.

 

It was interesting that the initial runs were cloudy with only 9 pounds of thrust, but after we got it running smoothly by adding less than 20% more screens, it was able to cleanly make 40 pounds of thrust at high pressure. Two possibilities are suggested, both with some supporting evidence:

 

The screens take a while to get “activated”. The end of the second run did start clearing up, which might have made the next run clear without adding any more screens. The tests on Saturday also seemed to warm up a lot quicker on the used pack. The fresh silver packs could gush raw peroxide a couple times before warming, but even after we had flushed the pack with water, the last run on Saturday was hot after only a short pulse. Some of that may be due to the different valve, but even a cracked open 1/2" ball valve probably flows a lot more than the solenoid.

 

A large part of the activity takes place towards the end of the pack at this size. When we took the motor apart on Saturday, the last quarter inch of screens had been thermally shrunk, and much of the pack above that looked relatively untouched.

 

 

Saturday tests:

 

media.armadilloaerospace.com/2002_01_29/saturday.xls

 

We did a lot of work on the test stand in preparation for biprop testing. The polycarbonate shield is now bolted to the stand around the engine, and there is a second, taller post so we can strap both tanks to the stand at proper heights. All plumbing that is used in the fuel side of things is brass, which we never use with peroxide, so we won’t accidentally mix them up. The intention is to start a normal monoprop run with the computer, and manually pulse open the fuel solenoid when it is running clear. We would only be doing this for a few seconds at a time, which the engine should be able to take in a heat-sink mode, because it has 0.6” of brass all around the chamber, and more around the nozzle.

 

We used gasoline today, but we are intending to use gaseous ethane as soon as we get a bottle of it. Working with a liquid fuel and another pressurization system is a pretty big hassle. A gaseous fuel that won’t puddle and self pressurizes will be a big advantage.

 

We replaced the solenoid with the ball valve for the tests today, and did most of our tests at low pressure, in preparation for biprop tests. After the great runs on Tuesday, we didn’t get a single clear run today. We did try pulsing some fuel into cloudy peroxide runs, but it didn’t light. When there is un-decomposed peroxide coming out of the engine, that means the chamber temperature is below the vaporization point, which is well below the autoignition temperature of the fuel.

 

1: 200 psi, 8/64” jet. Rough run, never cleared up. We were expecting the rough run, because the lack of significant changes in output with jet sizes on Tuesday implied that the solenoid was doing the metering, and that the jet was much larger than needed to flow the given results.

 

2: change to 5/64” jet. Still cloudy, but started smooth for a few seconds before turning rough. We are confused. Roughness at end possibly due to tank swirl that only shows up with ball valve?

 

3: close up a jet (with a hammer) so that it is even smaller than the 5/64” jet. Less thrust, still cloudy, smooth

 

4: I forgot what we tried here. Same results as 3.

 

5: higher pressure (400 psi). More power and still smooth, but still cloudy.

 

We took the engine apart to see if there was anything noticeable inside. The bottom 15 or so discs came out easily. We suspected that the motor might have stretched on the hot end like the little one we fired a couple weeks ago, but the dimensions were still correct. That means that screens had crunched themselves against the motor due to different expansion ratios. We pressed out the rest of the screens, and found that the anti-channel rings were no longer a tight fit either.

 

6: Reversed pack, putting the anti-channel rings and normal dimensioned screens at the end. It ran cleaner, but still not clear, and it was mildly rough. The roughness wasn’t too surprising, because we discarded the first five screens that were mashed up by the process of pressing the pack out of the engine, so the new pack probably wasn’t compressed as much as it should have been.

 

Our current theory is that the final long run on Tuesday ruined the fit on the anti-channel rings and the screens at the end of the pack, so all the runs today were channeling some peroxide around the outside.

We are going to try using gapless retaining rings as anti-channel rings next week, probably putting them through the entire pack, every ten screens. They will probably lose their spring after firing, but they should still be good fits.