|
|
Thread Tools | Display Modes |
#1
|
|||
|
|||
Unintended RCS Jet firing
An RCS jet has failed ON, not during a flight, but during a simulation. A
simulation that if it had occurred during a flight, would have resulted in an uncontrolled RCS jet firing. It happened four of five years ago during training for a flight. They were flying a TAL. At a point just after MECO, an RCS jet came on just as the flight software began an OPS transition. During the transition, some of the memory gets overwritten with the new code for the next OPS, and some of the the memory doesn't. There was a status word for the RCS jets that didn't get overwritten or reset, so the RCS jet that was ON before the transition remained ON after the transition. Essentially, the flight control didn't know that it was still commanding a jet ON. As you can imagine, the vehicle began to tumble out of control during a critical phase of a TAL. Critical because there isn't much time between MECO and Entry, and a lot of things have to happen in this short time span. As to what happened during the sim? I believe they stopped the sim to dump the computers to try to figure out what happened. If it had happened on the actual flight? I can imagine that the Commander would have figured out on her own how to fix the problem. It was similar to what happened during Gemini, where if something wasn't done very quickly, the vehicle would have continued to spin up. Craig Fink |
#2
|
|||
|
|||
Unintended RCS Jet firing
V-E-R-Y interesting, thanks for the report! I'm looking forward to hearing more details on this pretty amazing event. Frankly, I'm surprised, because the 'saw-tooth' command form was adopted in the early 1980s to preclude exactly this kind of failure. Can you estimate the year of the event a little more closely so I know who to ask for corroboration? "Craig Fink" wrote in message hlink.net... An RCS jet has failed ON, not during a flight, but during a simulation. A simulation that if it had occurred during a flight, would have resulted in an uncontrolled RCS jet firing. It happened four of five years ago during training for a flight. They were flying a TAL. At a point just after MECO, an RCS jet came on just as the flight software began an OPS transition. During the transition, some of the memory gets overwritten with the new code for the next OPS, and some of the the memory doesn't. There was a status word for the RCS jets that didn't get overwritten or reset, so the RCS jet that was ON before the transition remained ON after the transition. Essentially, the flight control didn't know that it was still commanding a jet ON. As you can imagine, the vehicle began to tumble out of control during a critical phase of a TAL. Critical because there isn't much time between MECO and Entry, and a lot of things have to happen in this short time span. As to what happened during the sim? I believe they stopped the sim to dump the computers to try to figure out what happened. If it had happened on the actual flight? I can imagine that the Commander would have figured out on her own how to fix the problem. It was similar to what happened during Gemini, where if something wasn't done very quickly, the vehicle would have continued to spin up. Craig Fink |
#3
|
|||
|
|||
Unintended RCS Jet firing
Hi James,
I could be wrong, but I'm pretty sure it was just prior to Eileen Collins' first flight as Commander (STS-93) and she was the one who found the bug. It had been there many many years. It just took the right set of circumstances for it to be found. Just goes to show that all the simulations that NASA performs on the flight software is really worth it. It also show, that no matter how much testing and verification is done, there can always be some unique path through the software that can cause a problem. A problem that may require some human intervention. If the computer is still commanding a 'saw-tooth' the RCS jet will think it's a valid command, it would just protect more of a wiring failure. A short "ON", not a bad command coming from the computer. Craig Fink James Oberg wrote: V-E-R-Y interesting, thanks for the report! I'm looking forward to hearing more details on this pretty amazing event. Frankly, I'm surprised, because the 'saw-tooth' command form was adopted in the early 1980s to preclude exactly this kind of failure. Can you estimate the year of the event a little more closely so I know who to ask for corroboration? "Craig Fink" wrote in message hlink.net... An RCS jet has failed ON, not during a flight, but during a simulation. A simulation that if it had occurred during a flight, would have resulted in an uncontrolled RCS jet firing. It happened four of five years ago during training for a flight. They were flying a TAL. At a point just after MECO, an RCS jet came on just as the flight software began an OPS transition. During the transition, some of the memory gets overwritten with the new code for the next OPS, and some of the the memory doesn't. There was a status word for the RCS jets that didn't get overwritten or reset, so the RCS jet that was ON before the transition remained ON after the transition. Essentially, the flight control didn't know that it was still commanding a jet ON. As you can imagine, the vehicle began to tumble out of control during a critical phase of a TAL. Critical because there isn't much time between MECO and Entry, and a lot of things have to happen in this short time span. As to what happened during the sim? I believe they stopped the sim to dump the computers to try to figure out what happened. If it had happened on the actual flight? I can imagine that the Commander would have figured out on her own how to fix the problem. It was similar to what happened during Gemini, where if something wasn't done very quickly, the vehicle would have continued to spin up. Craig Fink |
#4
|
|||
|
|||
Unintended RCS Jet firing
Jon Berndt wrote:
"Craig Fink" wrote in message An RCS jet has failed ON, not during a flight, but during a simulation. A Which simulator? I'm pretty sure it was the SMS. I wasn't involve with the simulation, just verifing the fix. Craig Fink |
#5
|
|||
|
|||
Unintended RCS Jet firing
Craig Fink wrote in message thlink.net...
Hi James, I could be wrong, but I'm pretty sure it was just prior to Eileen Collins' first flight as Commander (STS-93) and she was the one who found the bug. It had been there many many years. It just took the right set of circumstances for it to be found. Just goes to show that all the simulations that NASA performs on the flight software is really worth it. It also show, that no matter how much testing and verification is done, there can always be some unique path through the software that can cause a problem. A problem that may require some human intervention. If the computer is still commanding a 'saw-tooth' the RCS jet will think it's a valid command, it would just protect more of a wiring failure. A short "ON", not a bad command coming from the computer. Craig Fink James Oberg wrote: V-E-R-Y interesting, thanks for the report! I'm looking forward to hearing more details on this pretty amazing event. Frankly, I'm surprised, because the 'saw-tooth' command form was adopted in the early 1980s to preclude exactly this kind of failure. Can you estimate the year of the event a little more closely so I know who to ask for corroboration? "Craig Fink" wrote in message hlink.net... An RCS jet has failed ON, not during a flight, but during a simulation. A simulation that if it had occurred during a flight, would have resulted in an uncontrolled RCS jet firing. It happened four of five years ago during training for a flight. They were flying a TAL. At a point just after MECO, an RCS jet came on just as the flight software began an OPS transition. During the transition, some of the memory gets overwritten with the new code for the next OPS, and some of the the memory doesn't. There was a status word for the RCS jets that didn't get overwritten or reset, so the RCS jet that was ON before the transition remained ON after the transition. Essentially, the flight control didn't know that it was still commanding a jet ON. As you can imagine, the vehicle began to tumble out of control during a critical phase of a TAL. Critical because there isn't much time between MECO and Entry, and a lot of things have to happen in this short time span. As to what happened during the sim? I believe they stopped the sim to dump the computers to try to figure out what happened. If it had happened on the actual flight? I can imagine that the Commander would have figured out on her own how to fix the problem. It was similar to what happened during Gemini, where if something wasn't done very quickly, the vehicle would have continued to spin up. Craig Fink it can be done with hardware, too. www.gusgrissom.com |
#6
|
|||
|
|||
Unintended RCS Jet firing
wrote in message
om... it can be done with hardware, too. www.gusgrissom.com In the case of Mission 51-L, the firings observed from lift-off through t+58 seconds were discrete (on-off, bang-bang). Even the continuous, right-aft, yaw firings (beginning at about t+59 seconds) stopped briefly prior to the terminal firings. -- John Thomas Maxson, Retired Engineer (Aerospace) Author, The Betrayal of Mission 51-L (www.mission51l.com) |
#7
|
|||
|
|||
Unintended RCS Jet firing
"John Maxson" wrote in message ... wrote in message om... it can be done with hardware, too. www.gusgrissom.com In the case of Mission 51-L, the firings observed from lift-off through t+58 seconds were discrete (on-off, bang-bang). Even the continuous, right-aft, yaw firings (beginning at about t+59 seconds) stopped briefly prior to the terminal firings. -- John Thomas Maxson, Retired Engineer (Aerospace) Author, The Betrayal of Mission 51-L (www.mission51l.com) |
#8
|
|||
|
|||
Unintended RCS Jet firing
"John Maxson" wrote:
In the case of Mission 51-L, the firings observed from lift-off through t+58 seconds were discrete (on-off, bang-bang). Please provide telemetric evidence of these "observed firings". Even the continuous, right-aft, yaw firings (beginning at about t+59 seconds) stopped briefly prior to the terminal firings. Please provide telemetric evidence of these "right-aft, yaw firings". Please define "terminal firings". Evidence and definitions will provide a common ground on which we can discuss your claims. Roger -- Roger Balettie former Flight Dynamics Officer Space Shuttle Mission Control http://www.balettie.com/ |
#9
|
|||
|
|||
Unintended RCS Jet firing
"Roger Balettie" wrote in message m...
"John Maxson" wrote: In the case of Mission 51-L, the firings observed from lift-off through t+58 seconds were discrete (on-off, bang-bang). Please provide telemetric evidence of these "observed firings". Even the continuous, right-aft, yaw firings (beginning at about t+59 seconds) stopped briefly prior to the terminal firings. Please provide telemetric evidence of these "right-aft, yaw firings". Please define "terminal firings". Evidence and definitions will provide a common ground on which we can discuss your claims. Roger Roger, please comment on the S-11 rcs A&C roll switch. show us that you know something about flight dynamics. |
#10
|
|||
|
|||
Unintended RCS Jet firing
"James Oberg" wrote in message news | | What's "S-11"? At least, give us a clue which decade -- maybe, which space | program -- you're referring to? Jim, he's talking about Switch 11 on Panel 8 of the Block I command module in which his father unfortunately perished. Scott claims that switch was intentionally sabotaged with a small piece of metal, creating a short circuit which caused equipment to intentionally overheat and start the Apollo 1 fire. Needless to say, none of his claims stands up to much scrutiny. -- | The universe is not required to conform | Jay Windley to the expectations of the ignorant. | webmaster @ clavius.org |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Unintended RCS Jet firing | Craig Fink | Space Shuttle | 20 | August 17th 03 05:56 PM |