Re: [twuug] Space Shuttle Columbia blows up on reentry!

[Dave hamby <Dave.Hamby@xxxxxxx>]

Back in the mid-80's ACM Communications had a nice article on the  
Shuttle Flight Software. My recollection was that everything was in 4  
channels from sensor to actuator. IBM Federal Systems developed the  
computers and flight software for the application. The idea was that if  
a channel went bad, the happy machines would out-muscle the sick one  
and the bird would fly correctly. My recollection is that the flight  
software is customized for each mission.

On several occasions, one channel failed in flight, usually hardware  
but the vehicle was able to complete the mission and return home.

There was a fifth computer on board of independent manufacture and  
programming (Rockwell, I believe)  that could be used to reenter and  
land.

The article described the validation and verification procedure used.  
The Shuttle flight software is probably the most carefully tested code  
on the planet.  They were very careful about correctness and  
reliability.

Columbia had just been through a major overhaul or upgrade including  
new glass cockpits. This is a rolling upgrade and it is my  
understanding that the code has flown on other missions on other  
vehicles. I have no idea how much hardware or code was replaced. My  
understanding was that the flight deck display devices and software  
were updated to have functionality similar to that in a jet transport.

NASA is still reviewing telemetry and piecing together a sequence of  
high temperature indications and instrument failures on the left side  
of the vehicle.  Still too early for a hypothesis confirmation. One is  
that shed debris (insulation or ice) from the liquid fuel tank damaged  
the insulation of the left side of the vehicle. This came from post  
launch review of video footage. The flight rules check the ice build-up  
and won't allow the flight if ice accumulates to the point were the  
windows or insulation would be threatened if  ice were shed during the  
ascent.

The press conference said that the film of the ascent had been  
reviewed. Some debris were shed and struck the left side of the  
vehicle. Whatever happened, there were no indications of damage until  
the vehicle had begun reentry and was over California. From an analysis  
of the film (frame by frame) NASA concluded that things were OK and  
that there was nothing they could do differently  if there had been  
significant damage to the insulating tiles or wing. The reentry  
attitude and trajectory already minimize the heat and stress absorbed  
by the vehicle. It couldn't be made lower.

I'm somewhat surprised that they did not try to inspect the vehicle  
using the TV on the remote manipulator arm. One would think they could  
look at the upper surfaces but the external tank and are on the bottom  
of the vehicle and out of reach. The presenters mentioned that they had  
tried to photograph a shuttle in orbit  once before but that the  
results were of too low a quality to be useful.

WHRV has been carrying the NASA press conferences. One ran at around  
1700 Sunday that gave good detail.  Can't find it at NPR.ORG or  
NASA.GOV.

Dave

On Sunday, February 2, 2003, at 05:27  PM, Anthony VanCampen wrote:

>     I got a call from my sister in-law, who wanted to know if it was
> possible to reprogram the flight control computers on the shuttle.  My
> recollection is that there are three independent computers running the  
> same
> program and that they are dedicated to the flight management tasks.  I  
> old
> her that anything was possible, but that I didn't think that it was  
> likely
>     Does any one know what OS the flight comp(s) use?  My guess is  
> that it
> is an embeded OS with a very limited set of functions?
>
> Tony
>
>
> -
> ----------------------------------------------------------------------- 
> ---
> * To unsubscribe from the TWUUG discussion list, either send e-mail to
> -  twuug-request@twuug.org with the word "unsubscribe" by itself
> -  in the body of the message or visit:
> -  http://www.twuug.org/lists/twuuglists.html
>


-
--------------------------------------------------------------------------
* To unsubscribe from the TWUUG discussion list, either send e-mail to
-  twuug-request@twuug.org with the word "unsubscribe" by itself
-  in the body of the message or visit: 
-  http://www.twuug.org/lists/twuuglists.html