snidely@inteloa.intel.com (David P. Schneider) (05/02/89)
Atlantis and Magellan missed their first launch attempt by 31 seconds,
scrubbed when an anomaly was noticed appearently relating to a SSME fuel
(recirculating) pump.
I understand the pump is being replaced today, with launch Thursday or Fri-
day a possibility. (I'd rather they went for Saturday, but that's just for
viewer convenience).
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), and how big a hold can be dealt with at T -
31 seconds with little or no rollback of the count?
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.
At least one account I heard was that this was a very smooth countdown up
until T - 31 seconds, best since return-to-flight. NASA select, though,
had extra polling going on regarding a range safety computer that needed to
brought back on line. The boot sequence must not have been as rough as
some other issues, though, enven though it had the potential to scrub the
launch.
David P. Schneider
BiiN (tm)
Monday, 5.1phil@titan.rice.edu (William LeFebvre) (05/03/89)
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>
rbk@hpctdke.HP.COM (Richard Katz) (05/03/89)
-/ hpctdke:sci.space.shuttle / snidely@inteloa.intel.com (David P. Schneider) / 3:01 pm May 1, 1989 /
-At least one account I heard was that this was a very smooth countdown up
**********************
-until T - 31 seconds, best since return-to-flight. NASA select, though,
-had extra polling going on regarding a range safety computer that needed to
-brought back on line. The boot sequence must not have been as rough as
-some other issues, though, enven though it had the potential to scrub the
-launch.
I don't have the paper in front of me but it didn't seem that smooth.
Perhaps it was _relatively_ smooth - :-)
- range safety computer problems, as mentioned above
- overloaded receiver on Magellan
- recirculating pump
- blew a fuel line
- atmospheric pressure sensor in cabin failed
- two other problems which I can't remember off hand - if you
are interested, let me know and I'll post or mail it.
By the way, did anybody hear yet where the "metal particles" found in
the recirculating pump came from? Or where else they might have gone
and the effects? Or what caused the fuel line to spring a leak?
rich katz
hewlett packard
p o box 7050
colorado springs, co
80933-7050
email: rbk@hpctdlb.hp.comphil@titan.rice.edu (William LeFebvre) (05/05/89)
In article <4260012@hpctdke.HP.COM> rbk@hpctdke.HP.COM (Richard Katz) writes: >-/ hpctdke:sci.space.shuttle / snidely@inteloa.intel.com (David P. Schneider) / 3:01 pm May 1, 1989 / >-At least one account I heard was that this was a very smooth countdown up > ********************** >-until T - 31 seconds... >I don't have the paper in front of me but it didn't seem that smooth. >Perhaps it was _relatively_ smooth - :-) It was exceptionally smooth until the last few hours. There were no unexpected problems and no reasons at all to extend any of the planned holds. In fact, there wasn't even a threat of extending any of those holds, until those things that you mentioned started happening. That was all within the last hour or so of the countdown. William LeFebvre Department of Computer Science Rice University <phil@Rice.edu>
jmsc@inesc.UUCP (Miguel Casteleiro) (05/06/89)
In article <3223@kalliope.rice.edu>, phil@titan.rice.edu (William LeFebvre) writes: [ stuff deleted about how smooth was the countdown ] > It was exceptionally smooth until the last few hours. > [...] In fact, there wasn't even a threat of extending any of those > holds, until those things that you mentioned started happening. That was > all within the last hour or so of the countdown. ^^^^^^^^^^^^^^^^^^^^^^^^^^ Yeah, that's when the going gets tough!!! -- __ Miguel Casteleiro at __ /// INESC, Lisboa, Portugal. \\\/// Only UUCP: ...!mcvax!inesc!jmsc "Life is hard and then you die." \XX/ Amiga
fosler@inmet.UUCP (05/13/89)
From AvLeak, One of the two redundant range safety Cyborg computers failed at 38 min. prior to scheduled launch. Flight rules require that both the primary and backup systems be on line at liftoff. The computer was brought back on line in 18 min. If the computer had not gone down, launch would have occurred before the recirculation pump shorted. Recirulation pumps are activated 30 min. after start of the liquid hydrogen fast-fill cycle, when propellants are loaded in the external tank. They had been operationg 7hr. 45 min. when they shut down. The pump shorted because a sliver of metal contacted the exposed part of a pin in the connnector between power supply and the pump. NASA does not know the origin of the metal particles, which were solder, aluminum and steel. There is some belief that they came from the pump itself. Carl Fosler