May 23, 2001 Meeting Notes

 

In attendance:

 

John Carmack

Phil Eaton

Russ Blink

 

New supplies:

 

Servo operated ½” ball valve

Larger electronics box

Mil-spec connectors

New 12v battery and charger for electronics box

More Tefzel wire

 

On order:

 

3 KVH Fiber Optic Gyros

Two liter flasks

-10 hose and fittings for test stand

 

Engine Testing

 

We did engine testing on the small motor today with very interesting results.

 

media.armadilloaerospace.com/2001_05_23/smallMotor.xls

 

We are planning on using roughly 50 pound thrust engines four the four attitude control engines on the manned vehicle, mostly because that is about the flow limit of the NOS Super Pro Shot solenoids as determined by firing it on Juan’s 100 pound thrust motor. The Super Big Shot flows more, but we want to avoid using those 30 amp solenoids if possible.

 

The small test engine makes about 13 pounds of thrust in its last tested configuration, with a 0.25” diameter throat. We wanted to see if the same size motors could still be used with a 0.5” diameter throat, which should give us about the thrust we want. That would defiantly be pushing it, because the inside diameter of the motors is only 1.1” diameter, and going below a 4x area ratio between combustion chamber and throat starts bringing in losses.

 

The question was if we were going to be able to fully catalyze four times the flow we normally use. We started by loading the engine with one of our old catalyst packs that was about four times what we wound up eventually cutting them down to.

 

Russ drilled the nozzle out to 0.375” where it still had a tiny expansion lip, and we loaded it with 250 ml of peroxide. It catalyzed perfectly, but it only made about 16 pounds of thrust, which was a bit less than we expected even without the nozzle, and it had a buzz in the thrust curve.

 

We then drilled it out to 0.5”, which is now just a hole in the end of the motor.

media.armadilloaerospace.com/2001_05_23/nozzleless.jpg

(we are using the 30 amp Super Big Shot solenoid on the test stand – it is probably twice as heavy as the engine…)

 

Phil didn’t think it was going to make much more thrust, theorizing that it was flow limited through the catalyst pack. He was right – it didn’t make any more thrust, and it still had the buzz.

 

We then cut the catalyst pack in half, and run it again. We ran out of nitrogen during the pressurization, so we only got 375 psi (instead of 430), but the run still made 23 pounds. The buzz was also gone, but we couldn’t tell if that was due to the lower pressure or the change it catalyst back pressure.

 

Rather than cutting the pack down some more, we tried out what we have theorized a bit about in the past: we took four of the unplated nickel foam discs, compressed them four to one, and put that on top of fifteen (the same number we have been using in the VTVL engines) completely uncompressed silver plated foam discs. The idea is that the highly compressed nickel discs will spread the peroxide and decouple the pressure feed system from the catalytic activity, then the peroxide can flow freely through the rest of the catalyst

 

It worked strikingly well. It made 40 pounds of thrust, and ran smooth.

 

Getting a 3 to 1 expansion exit cone on the nozzle should give a coefficient of thrust around 1.35, which should give us right around 50 pounds of thrust (we will drop a bit with the one size smaller solenoids).

 

It is interesting to note that the only changes between the last three runs are catalyst pack changes, giving 2.5x more thrust and getting rid of the buzz.

 

Russ is going to make four completely new nozzles with 0.5” throat and 3x expansion ratio, which we will be able to just screw on the existing engines. With the current plumbing, they will probably only make about 25 pounds of thrust, but that will still give the VTVL the ability to accelerate at more than 1G straight up. When we replace the plumbing with the same stuff we will use on the bigger vehicle, it should be able to pull 3Gs.

 

Once we get gyros that aren’t G sensitive, we will install some fins and parachutes on the VTVL and really let it fly. Eventually we will want to do a complete powered landing, but it is going to be extremely cool just to lift it off the ground and hover for a second, then tear off at 3Gs…

 

Other Propulsion Issues

 

The small plastic solenoid we were testing at 150 psi turned out to have a nitrile seal, which is another form of Buna-N, which is nor peroxide compatible. I am still interested in looking at 150 psi valves for some applications, but I need to find a good (C)PVC valve with Teflon or Viton seals.

 

Neil brought a 32 liter NGV tank that we are considering for the manned vehicle. It is fiber wrapped, but mostly it is a heavy hunk of aluminum. It weighs 60 pounds, and is rated for 3000 psi, which is huge overkill for us, but the extra mass still won’t be as dominant as the pilot’s mass. We need to get it well cleaned, and I need to buy some adapters to plumb it up.

 

Having ports on both the top and bottom is going to be a nice improvement for us. We can load peroxide with a normal pump (or just a funnel) instead of drawing a vacuum in the tank, and we won’t have to blow the nitrogen up through the peroxide when pressurizing. We will also be able to monitor gas temperature and pressure, allowing us to make a fairly accurate propellant gauge.

 

The kzco servo ball valve seems to be a pretty nice unit. It was $395 for a stainless/Teflon ½” ball valve with servo motor and potentiometer feedback. It masses 1.7 kg, but that will hardly register on the next vehicle. Ball valves up to one inch can use the same servo system, and they have larger ones for valves up to several inches. We are probably going to have to use four solid state relays to allow switching the current direction for forward / backwards, and I am going to need to write some potentiometer closed loop control code to run it from the remote joystick.

 

The main lifting engine will probably have about a 1.75” diameter throat (Russ: note that I increased this), which should give something over 500 pounds of thrust. The ½” servo ball valve is probably good for about 2000 pounds of thrust if we ever max it out.

 

 

Software

 

I wrote some graphing software for our flight logs, and gathered together the first three sets of tests:

 

www.armadilloaerospace.com/misc/FlightLogs.htm

 

I have updated the 3D flight simulator to model blow down pressure changes (but not temperature changes) and use roughly the correct values for the current vehicle. I need to get the updated version put on the web site.

 

As soon as we get higher bandwidth gyros, changing to a continuously variable attitude control system should give us large benefits in stability, control authority, and response time.

 

Gyros

 

The specs on the WatsonGyro 3 axis unit are:

 

 ARS-E332-2A is  $3408.

This triaxial gyro vibration survival 10G rms (20 to 2Khz.

Delivery 3 weeks ARO.

 

 

Heath Albritton got a quote from Systron Donnor for their angular rate sensors:

 

>The AQRS rate sensor is available in two variations, -104 and -109.  The

>only difference between these two units is unit price and Bias

>performance vs. temperature.  Budgetary pricing for the AQRS rate sensor

>in quantities of one to three units is $600 for the -104 variation and

>$300 for the -109 variation.  A mating connector is included with each

>shipped unit.  Please note that SDID is currently only manufacturing the

>AQRS in the 75 deg./sec. full range.

 

The specifications can be found here:

http://www.systron.com/prodinfo/AQRS.html

 

 

I also got some more info from BAE systems, but they didn’t provide a price.

 

 

I decided to order the KVH fiber optic gyros, at $1500 / axis. We are going to have to make up a fairly precise set of perpendicular mounting plates for them. Having a full 100hz bandwidth and no shock, G, or cross axis sensitivity at all will be very nice.

 

 

Auto Landing

 

I recently talked to a few of the people that worked on DC-X, which was interesting.

 

DC-X used a Honeywell radar altimeter from a helicopter, but it was blended as one of several inputs into their guidance system. Bogus data from it at one point was the cause of the harder-than-expected landing on one of the DC-X flights.

 

The radar unit was pretty bulky, and I was urged to purse a GPS / inertial based solution.

 

I will probably start working some on this in the next week or two. I don’t plan on going all the way to a Kalman filter for blending, but instead plan on just buffering accelerometer data, and back-correcting to the point of the GPS second pulse after all the GPS data has been read in over the serial line. I will probably just use Z axis acceleration, because inside of one second, in the attitudes we will be landing at, the cosine variability won’t be very large.

 

Tim Nolan volunteered to build a rangefinder for us to try. The initial one will be ultrasound, which I am rather dubious of working next to the rocket engines, but he has also found some low cost radar parts. I still have a preference for a “direct reading” type of sensor like this, but we will see how the GPS + accelerometer works out.

 

It looks like our project will probably be the first to demonstrate vertical landing with an engine restart after ballistic flight, which is still an often sited problem with VTVL. The DC-X, and I presume the Japanese VTVL project, don’t have easily restarted engines, so they have been constant thrust only.

 

 

 

Manned Vehicle Design Issues

 

Relief valve vs burst disk

 

Piloted solenoid vs servo ball valve for purge valve. Probably need a ball valve for high flow, and because it might be called upon to operate at over 1000 psi if the tank is cooking off.

 

Do we need check valves in front of engines? The half inch check valve will be a LOT more restrictive than the half inch ball valve, and it may prevent us from hitting our thrust targets. I did realize that the only valid place for the check valves is between the controlling valves and the engines, because any other location would leave trapped peroxide in the system in the event of a purge.

 

We probably need another servo ball valve for the master cutoff valve, although in the event of an emergency, purging might be almost as effective.

 

 

Misc

 

My wife is closing on 100 acres of land in Rockwall next month, so we should have plenty of room for a noisy static test stand and some moderate altitude flights.

 

We are going to have Armadillo Aerospace shirts soon.