Saturday, December 21, 2019

Bo-Bo-ing - A case of no physics

I never write the names right because of the history-destroying search engines of nusa.  But this is a fine example of new engineering without physics.

When sending off something like a capsule, or an airplane, you should have a full physics model, or don't do it.  They didn't know the physics of the plane, and now the capsule.  And nusa falls right in there.

So, after a massive reading of articles I can't find any more, I go through the failure:

Up in orbit, the capsule fired a thruster.  This was the main engineering failure because the thruster didn't light, and it was mindlessly thinking it was firing.  This slipped the capsule way off course.  If you have a full physics model, then you know what that thruster is going to do in terms of rotation and acceleration.  You measure that with simple accelerometers.

This is the most fundamental error for automated systems -- you must always have feedback.  An ATM gives out money, but a physics model measures the bills going out.  Otherwise you didn't pay the guy and think you did, or the machine spews out tons of bills.  They made that mistake a lot at the beginning.

The rest of the capsule action was a comedy of errors making up for the first one.  Finally, the capsule went into a panic state, firing all thrusters continuously to stay in one position.

Nussie tried to paper it over by saying that, had astronauts been in there they would have corrected sooner.  Huh?  People should go into this capsule-max?  People don't do rapid course corrections.  They are automated for a reason.  Was there a manual for correcting a false reading?  This is all looking horribly familiar.

There must be some heavy corruption here, more likely technical corruption, like diesels-gate.

No comments: