🚀 go-pugleaf

This is all from memory, so forgive me if there are a few minor errors.


In article <4361@omepd.UUCP> snidely@inteloa.UUCP (David P. Schneider) writes:
>Atlantis and Magellan missed their first  launch  attempt  by  31
>seconds...

This is something that needs to be cleared up.  This "only 31 seconds"
jazz is getting to me....  I'm not flaming Mr. Schneider, but I've heard
the news people say "with JUST 31 seconds....." as if to say "they came
SOOOO close...."  So I'll take this opportunity to speak my mind.

The problem was actually noticed at about T-1:15, but the hold didn't take
place until T-31secs.  They have planned times for holds:  T-20min,
T-9min, T-5min, and T-31sec as well as others.  They only go into a hold
at those times (although I suppose something real serious would make them
stop the clock).  If anything happens between T-5min and T-31sec that
violates launch criteria, they will continue the count anyway until the
T-31sec hold.  Then they go into a hold.

Why T-31 seconds?  Because it is at that point that the automatic launch
sequencer takes over and then everything is handled directly by computer,
effectively controlled by the on-board computers (GPCs).

>Some questions about this: is this one of  the  pumps  that  was  installed
>after  Atlantis  was  on  the  pad,  how  was  the  anomaly  noticed, which
>desk/function/station announced the hold (it sounded like a woman's  voice,
>such  as had been calling off some checklist items a little before, if that
>helps with the identification)

Booster would have been the first to notice it, but the woman you heard
was probably some sort of overseer or coordinator.  That is, she was
probably not the person that first noticed the problem---she was probably
just relaying the call.  (I know a whole lot more about how mission
control/JSC works that I do launch control/KSC).

>and how big a hold can be dealt with at T - 31 seconds with little
>or no rollback of the count?

They cannot hold at T-31sec for more than a few minutes, because the
Auxiliary Power Units (APUs) are running.  Once they are started, launch
must happen within something like 15 minutes or they have to roll back
the count and recharge the APUs (I believe that takes hours, so it would
force them to scrub for the day).

>From the sound of things, once the hold was announced, having to scrub  was
>pretty  much  automatic  because the sequence could not be recovered within
>the launch window.

And because of the nature of the failure, even if they had tried again,
they would have failed.  Of course, they didn't know that at the time....

>At least one account I heard was that this was a very smooth  countdown  up
>until  T  -  31 seconds, best since return-to-flight.

I had heard that it was the smoothest countdown in the history of the
shuttle program, up until they held for the RSO's computer.

>NASA select, though,
>had extra polling going on regarding a range safety computer that needed to
>brought  back  on  line.

The range safety officer (RSO) had to reboot his computer at some point
during the morning.  That was why the launch was delayed about 5 minutes.
Since I didn't tune in until after T-9 minutes, I don't know at what point
in the countdown they held for that.  I didn't actually hear this
happening---I heard about it later.

>The boot sequence must not have been as rough as
>some other issues, though, enven though it had the potential to  scrub  the
>launch.

They will not launch if the RSO doesn't have his computer.  Even though
the computer is on the ground.

			William LeFebvre
			Department of Computer Science
			Rice University
			<phil@Rice.edu>