Monday, August 09, 2004

Two up, two down (in pieces).

This was not a good weekend for two X-Prize contenders.

Armadillo's failure is particularly disheartening. Their engine was showing problems during setup for a hovering test, catalyst rings fell out during setup, and the proceeded with the tests anyhow. (On the basis that such problems had occurred before, and had not caused a failure. Sound familiar?) After a 'successful' hovering test, they proceeded with a flight test on the next day. While warming the engine up as part of the flight test, the engine behaved in an anomalous fashion (it warmed up slowly and 'blew' raw propellant from the nozzle), and they proceeded with the test anyhow. During flight, problems were encountered with the engine's throttle controls (the write-up is unclear on whether or not those problems were related to the engine failure), and it ran out of fuel in-flight, resulting in a crash rather than the planned landing under thrust.

John lists some valuable lessons learned during the flight and subsequent failure, but somehow misses the big one; when you have a test anomaly, stop and find out why.
(Slashdot discussion of these problems here .)


At 1:38 PM, Anonymous Anonymous said...

After reading John's entry - in its entirety - and seeing a couple select pictures, there are a few points made about what went wrong. The mid-plate bent due to improper welding, which from the look of it deflected flow; it's mentioned that the catalyst likely settled while the vehicle was in transit on the trailer, and John mentions a probable fix for this; obviously, no "fuel gauge", an item certain to be integrated in the next test vehicle; John readily admits they should have stopped, hindsight being 20/20. There, now you know "why", and I saved you from having to read the whole thing. Cheers!

At 5:03 PM, Blogger Derek said...

I did the read the article and saw that John learned the why in a post-mortem, my point is that it should never have come to that. Twice he had failure symptoms and twice he ignored them. The fuel gauge isn't a fix, because fuel level wasn't the real problem. The problem was excessive (and visible and acknowledged) fuel consumption pre-launch.

At 10:07 PM, Blogger Jake said...

I was actually discussing this with my roommate on the way into work today. My roommate designs FPGAs and I write software, and both of us think that this is John's commercial software experience coming back to haunt him. When the cost to fix something is a quick recompile and maybe reboot, "this might work, let's give it a shot" is the rational way to go about things. When it takes thirty grand and five weeks for long lead time parts, that's not such a good idea.

At 4:15 PM, Blogger Derek said...

A further update on Space Transport's launch can be found here. It isn't pretty.


Post a Comment

<< Home