February 12 and 16, 2002 Meeting notes

 

In attendance:

 

John Carmack

Phil Eaton

Russ Blink

Joseph LaGrave (Tuesday)

 

Regulated Test Stand

 

On Tuesday, we did a lot of work modifying our test stand to allow pressure regulated biprop work. I got more AN swivel fittings, so we moved the jetting around so that the kerosene jet is right at the engine for high speed injection, and the peroxide jet is now also a flare jet instead of the hand-made NPT union jets we were using before. I didn’t want to spend time making a lot of little flare jets, so I went looking to see if nitrous jets went up to the larger sizes we would like. I was pleasantly surprised to find that Edelbrock offered a huge set of AN flare jets from 0.018 to 0.130, and they are made of stainless steel instead of brass like the NOS jets. I bought the “master tuner jet kit”, which has eight of each size. We are probably going to add jets on all the lander attitude engines the next time we test. With the new plumbing arrangement, we can change jets on the test stand in under a minute without even depressurizing our tanks.

 

We were a little concerned that the spare regulator we had wouldn’t flow enough to hold a constant pressure, but I logged ullage pressure today, and the runs only dropped one or two psi when they started, and held steady there.

 

Here is the new fuel injector ring that I made, which allows us to change nozzles and cat packs independently, and will allow us to swap a hybrid grain section in place of the fuel injector section. We test fired it with water, and with the jet right at the injection port, the water slams against the other side at very high velocity, atomizing well. An impinging jet injector would be almost impossible to align with tapped holes and held jets, and at our fuel flow rates, the required jet sizes would be smaller than we can get or make, so we will stick with single point injection for a while.

media.armadilloaerospace.com/2002_02_16/injector.JPG

 

Russ managed to get a water jacketed combustion chamber made this week. The internal chamber is 1” diameter and 2.75” long, which gives us well over the recommended L*. We hook one end up to a garden hose, and the other end up to some tubing that ejects in front of the test stand so we can see if we ever get any steam.

media.armadilloaerospace.com/2002_02_16/cooled.JPG

 

Here is the test stand before we connected the nitrogen bottle.

media.armadilloaerospace.com/feb16_02/stand.JPG

 

 

Test Runs

 

Matt, our camera guy, was out of town today.

 

run 1:

Unchanged monoprop from last Saturday, except the NPT 0.060” jet on the peroxide was changed to a 0.060" AN flare jet.

The pure silver screen pack warmed up on the very first pulse from stone cold, and continued to catalyze perfectly.

Still running rough, as it was last week.

 

run 2:

We removed the check valve from the peroxide side, theorizing that the check ball gave pressure pulsations more leverage.

This run was rougher than the last.

 

run 3:

Still with the check valve out, reduce to a 0.040 peroxide jet.

Less thrust, a little smoother, but still not very good.

 

run 4:

Change back to 0.060 jet, swap ball valve for solenoid, replace check valve.

Still rough.

 

run 5:

First hot fire of the regulated biprop. 0.024 fuel, 0.060 peroxide, 150 psi.

Did not light. Not unexpected, as our previous tests were with a higher peroxide tank pressure to compensate for the cat pack drop, so this was much richer.

 

run 6:

0.040 peroxide, 0.018 fuel

Still did not light.

 

run 7:

The 0.018 fuel jet is as small as we have, so increased the peroxide side to 0.050. Pressure increased to 200 psi.

It lit instantly on each press of the fuel button, and made lots of mach diamonds in a beautiful exhaust jet!

The roughness was worse in the hot firing sections.

 

run 8:

same jets, 250psi pressure

We tried to record this on my little camera, but the battery died literally two seconds before we hit the hot fire button.

Still rough.

 

run 9:

We made a new cat pack in the second 1” by 2” body that I made. The formula is currently: five layers of: ten each alternated 80 mesh silver screen and 20 mesh stainless screen with a gapless retaining ring ahead of each block for a total of fifty silver screens, fifty stainless screens, and five retaining rings. The remainder of the pack space is taken up with the retaining plate and a 0.25” spacer, and about 0.2” is compressed down during the tightening of the pack chamber onto the fuel injector. There is no spreading plate at the top.

The run was smooth, like the first 80 mesh pack when we first started running it, but not quite the absolute-dead-smooth of the (slow warming) 20 mesh pack we made a couple weeks ago that later crapped out on us (before we were using the gapless retaining rings). It will be interesting to see if this pack also gets rough after a dozen runs.

The hot fire sections were also smooth.

 

run 10:

We see a bit of smoke above the plume, and there is soot in the engine, so we believe it is still running rich. Increase the peroxide jet to 0.060.

Still a bit of smoke, but still smooth on both monoprop and hot periods. There was ten seconds of almost continuous hot fire with only a brief pause with the fuel off.

 

run 11:

Increased peroxide jet to 0.070.

This was still smooth on the monoprop periods, but rough on the hot periods.

 

run 12:

back to 0.060 peroxide, exactly as test 10.

Still rough on the hot periods. Strange.

 

run 13:

Increase fuel jet to 0.020.

Still rough on hot period, obviously richer.

 

Analysis

 

We still don’t have a handle on the rough monoprop running. About our only theory is that the 80 mesh screen may be interacting with the gas flow in some way. Our big loads of catalyst still aren’t here, but if we need to, we still have an engine’s worth of 20 mesh silver screen that we can reuse for tests.

 

We have a theory about the rough hot fire sections. Because we are using the smallest fuel jet, our thinking is that we don’t have sufficient pressure drop across it. The two runs where it was smooth may have had a more restricted jet due to soot or a foreign particle. Tiny nitrous jets are well known for getting clogged up. We can’t go up much on feed pressure because of the rating on the fuel tank, so Russ is going to bore out the nozzle from 0.25 to 0.40, which will increase our total mass flow enough to hopefully choke the small fuel jet. That will also reduce our L*, but there is a bit more meat that can be bored out of the chamber to compensate.

 

It is definitely possible to be to rich to light, even with a big chamber and good atomization, which places a limit on how much cooler you could get an engine to run overjetting. At optimum mixture ratios, peroxide / kerosene is hot enough to melt TZM, so if we want a completely passively cooled engine, we will need to do something to lower the chamber temperature (and give up some Isp).

 

The water cooling seems to work fine. After a hot fire, the exiting cooling water is only slightly warm. We might try taking the cooling exhaust out to where a person can monitor the temperature during firing.

 

The current cat pack formula seems very solid from a catalyzation standpoint.

 

Leaving fuel in the tank between runs is extremely convenient. However, when we tried to spray the remainder at the end of the day back into the kerosene can, we wound up making a mess after the last of the kerosene came out of the tank and the nitrogen started blowing by. Next time, we are probably going to leave the kerosene in the tank and just vent off the pressure.

 

We improved both the injector and the chamber volume over the runs last weekend, but we could tell that the chamber volume was the more important factor, because on some of the runs we got spatters of combustion after the fuel valve was closed, probably a result of it sucking the remaining little bit of fuel left between the check valve and the jet into the chamber. Even the non-atomized spatters had mach diamonds.

 

After we actually get some good video of the biprop running, we will probably take the test stand out to the 100 acres and see if the biprop can run steady state for 60+ seconds with water cooling (we will have to use another pressure vessel for the water, there isn’t city water out there). If that works out fine, we may try making a regeneratively cooled engine, possibly out of aluminum.

 

 

Other Stuff

 

The Oklahoma Space Industry Development Authority (OSIDA) meetings are continuing to go well. I have been driving up to Oklahoma City for the meetings, and we should be signing a memorandum of understanding soon. It is looking good for getting airspace for manned supersonic work late this year, working towards 100km shots next year.

 

The ethane arrived, but we need to get a special CGA fitting for it before we can try it out. We seem to be getting good combustion with liquid kerosene, but when we try building bigger biprops, we will probably want to start with gaseous ethane because it removes the requirement of getting good atomization and distribution. The metering will also be a lot more positive with a gaseous fuel, because the orifice will be larger, and it will be choked sonic flow.

 

The Laser Atlanta Advantage rangefinder arrived, and I have it configured as we will use it on the vehicle. It gives 240 samples a second, with an accuracy of a couple inches, and a range of 2000’ without a reflector (20,000’ with a reflector). We will be logging this data on the next lander flight, but we won’t be using it for auto-hover and auto-land until later tests. Mounting it is going to be a bit of a trick.

 

Joseph will soon have a 1000 square foot concrete pad poured for lander operations and high speed rotor testing at my 100 acre lot.

 

Our airplane parachute is supposed to be here on Tuesday.