April 18, 2004 notes

 

Better liftoff

 

After missing both the originally quoted and later verbal ship dates, the thermocouple amplifiers from Axiomatic finally arrived, and, wait for it… they were the wrong supply voltage.  I wasn’t about to wait for them to correct it, so I looked around until I found another supplier.  http://www.piresearch.com/ had just what I wanted, and they fed-exed three two me the next day, and said they could get more from the UK by Friday.  On Friday they called and said that they had been hung up in customs longer than expected, but they could ship it Saturday delivery if I wanted.  Quite a difference in service.  The Axiomatic parts went back for a refund.  The Pi parts are intended for auto racing exhaust temperature measurement, so they don’t include a precise calibration, but they are doing the job.

 

We built some metal stands that the vehicle could rest on during warmup and before liftoff, then have them just fall over so the vehicle would have free-swinging space under the crane, but it kept falling off of them during the warmup pulses when it lightened up a little.  I would really like to get this working so the vehicle doesn’t have to start out rotating and tipping, and I can give everything a really good warmup, but we may need to make them a bit more self supporting, and actually pull them out from under the vehicle on liftoff, but that is potentially error prone.

 

http://media.armadilloaerospace.com/2004_04_18/stands.jpg

 

The control gains were doubled for the test this week, and we got a little more tank pressure in it.  At hanging liftoff there was already 10 degrees of roll, and a few degrees of yaw.  I’m still not giving it a good enough warmup.  It lifted off fairly straight, but it didn’t stop rising as quickly as I wanted it to.  I clicked for it to come down a bit, but it also picked up five degrees of tilt in the first couple seconds.  It was able to correct for the tilt, but it came up taught on the tether, which jerked it back past the 20 degree tilt abort point, causing the computer to shut it down.  It doesn’t look like it would have had a problem continuing the flight if the tether wasn’t there.  Having demonstrating the automated aborts over and over is probably going to be a helpful thing when have to discuss vehicle safety with the regulatory bodies.

 

After we steadied it down, we got ready to hop it back up with the remaining propellant, but just as it started throttling up, we lost the computer.  Everything shut down fine, on the crash, another demonstration of good aborts.  The computer came back up quickly, showing that it had been a momentary power interruption of some kind, because we do not run a watchdog to automatically reboot the computer in the event of a hang.  However, none of the serial devices were responding, and the A/D values were all off scale low.  When we did a manual power cycle, which brought the serial devices back, but the A/D board is still dead.  This is the second board to die recently.  We have reason to believe that the problem is due to the spark plug ignition system.  We have caused our dataq system to freak out on the test stand several times, and the blown A/D boards have only happened since we had the spark ignition systems, never with the glow plugs.

 

We moved the ignition coils to keep the plug wires farther away from the thermocouple wires, and we will probably add some metal braid around them for added shielding.  I would like to investigate using a coil-on-plug system to eliminate the wires completely.  Does anyone have any experience using particular OEM COP systems for custom applications?  We are going to add protection circuits and possible linear opto isolators to the next custom breakout board we make.

 

Some of our wrapped insulation tape was starting to fray from the engine blasts now that it has been hot enough that the initial sticky-backing was cooked off.  A few more hose clamps should keep it all properly in place.

 

From the telemetry I was able to tell the following:

 

The reason the vehicle continued rising longer than expected was because the throttle was clamped at my minimum for stabilized flight, which was 25%.  I have that limit because I don’t want an engine to shut completely off during any correction, but at the higher tank pressure we used today, 25% throttle was only –1 m/s acceleration, so it was barely slowing down.  If we had 30 psi more in the tank, the minimum throttle would have been actually accelerating, and I would have had to kill the thrust completely instead of commanding a down velocity.  I am going to change the minimum throttle to 15%, which is just barely cracked open.  I was also using this value for the stabilization level after boosting for accelerating flights, but it looks like I will probably have to use a minimum vehicle acceleration as the coast throttle limiter, because a fixed throttle percentage is going to vary too much with these ball valves at different pressures.

 

I am going to increase the gains some more for the next test.  The current values were sufficient to let the vehicle make complete back-and-forth corrections in all axis, but the period and amplitude were both larger than I want.  The current values are probably acceptable with full size engines, which would have a lot more authority for a given differential throttle.

 

The A/D board had a problem right after the first warmup pulse (the first time the spark ignition was activated).  The tank pressure transducer went from ready 210 psi before starting, fluctuated around a lot during the pulse, then stayed at 146 psi for a long time, before gradually creeping back up to 210.  The valve pot feedbacks, thermocouples, and chamber pressure transducers did not show any problems.  In the past, we did see a pressure transducer (different brand that we no longer use) short internally and draw enough current to smoke things, but I still feel the spark is more likely the source of the problems.

 

http://media.armadilloaerospace.com/2004_04_18/bigHover2.mpg

 

 

Jet vane vehicle

 

We cut off the top of the test engine and pulled the spreading plate out of it, then brazed one with less flow area in while under 2000 pounds of pressure.  The engine runs great now, but the fact that we changed two things doesn’t allow us to definitively say what the cure was.  I think it was the braze-under-pressure.  We finally realized how silly it was to have our press on the complete opposite side of the shop from our cutting / welding / machining tools, and swapped some things around for better working conditions.  There have been several cases like that over the years where we realize that things just weren’t organized intelligently.

 

The jet vane vehicle is almost ready to fly.  We just need to add a heat shield for the actuators, and build a protective cover for the electronics.  I am really looking forward to flying this and contrasting the issues with the differentially throttled vehicle.  The flight control software just takes a switch to determine which vehicle it is going to fly.

 

http://media.armadilloaerospace.com/2004_04_18/vaneTestFit.jpg

http://media.armadilloaerospace.com/2004_04_18/engineOverVanes.jpg