STS-27 didn't involve an abort but it's worth reading about as well due to its direct relationship to the Columbia disaster: https://en.wikipedia.org/wiki/STS-27
Just two missions after Challenger, the orbiter experienced tile damage much like Columbia did 15 years later. It left an entire field of damage all the way down one side of the orbiter. They inspected under the orbiter using the canadarm as much as they could, and found some early indications of tile damage. Mission Control refused to use a high-resolution unencrypted camera to image under the orbiter due to the classified nature of the mission payload, instead using a lower resolution encrypted link, and dismissed the damage as video artifacts.
The crew wasn't happy at all at this assessment and seriously believed they were going to die on reentry. Mike Mullane believed he would see hot aluminum streak up the windshield like raindrops. Commander Hoot Gibson advised the crew there was no point in "dying all tensed up". Even if Mission Control had deemed this to be inevitable, it was not at all clear how quickly or even if NASA could process another orbiter and evacuate them in time.
The shuttle's wings have a combined elevator-aileron called an "elevon"; when both are in the same direction, they focus as an elevator (controlling pitch), and if they are in opposite directions, they function as an aileron (controlling roll). On reentry, Gibson kept his eyes glued to an elevon gauge to see if aerodynamic forces from a disintegrating airframe were forcing the onboard computer to roll the shuttle, creating a difference in degrees across the elevons. If they had deviated more than a few degrees, he planned on cussing out NASA before he and the rest of the crew died.
Fortunately, the shuttle made it back safely, but the damage was far worse than anyone anticipated. The foam strikes had spared the RCC leading edge panels that would eventually doom Columbia, however, nearly one half of the bottom of the shuttle had some degree of tile damage, and one had gone missing completely in a critical section. That section, by the grace of god, had a heavier plate welded to the airframe that was used as part of a communications antenna. One tile missing anywhere else, and the shuttle would have been destroyed.
Neither STS disaster came out of the blue. Each had earlier clues that were ignored, and that's what both accident reports highlighted.
I think STS had so many high-criticality problems that there was probably no way to salvage it as a vehicle. The only sad thing is that they didn't sunset it after Challenger.
My "favorite" Space Shuttle abort mode is the Return to Launch Site (RTLS) abort[0] because of the complex energy management procedure necessary to pull it off.
As I understand it the energy management of an RTLS is tenuous because, ultimately, the orbiter is an unpowered glider (with a glide ratio similar to a brick). The limits of stress on the orbiter's airframe are a factor, too.
The orbiter's mass, energy (altitude), and downrange position all have to be in a reasonably tight envelope to make the landing without coming up short or overshooting (there's only so much energy you can bleed off doing S turns).
John Young[1] described an RTLS abort working thru "continuous miracles, interspersed with acts of God".
Somebody made a fun (if you like this kind of stuff) simulation of an RTLS abort[2] in Kerbal Space Program.
This mission also illustrates how confusing the old STS mission numbering scheme used to be.
STS-51-F is the only exercised launch abort, and the Challenger was the orbital vehicle.
STS-51-L, which occurred the following year, is the famous Challenger disaster.
Originally, the numeral in the mission designation indicated the federal fiscal year of the mission and the launch site (51 being 1985 and launch site 1 (KSC, 2 would have been used for Vandenberg). The letter indicated the sequential launch index, so A=1m, B=2, etc.
This was originally done after STS-12. Superstition around Apollo 13 created the desire to avoid an STS-13. The numbering system returned to a strictly sequential schema after STS-51-L, with STS-26.
> Booster Systems Engineer Jenny M. Howard acted quickly to recommend that the crew inhibit any further automatic RS-25 shutdowns based on readings from the remaining sensors,[6] preventing the potential shutdown of a second engine and a possible abort mode that may have resulted in the loss of crew and vehicle (LOCV)
Incredible that human action during ascent (when everything is usually automated) potentially saved a crew. One of probably dozens of cases, is my guess.
STS still has the record for most astronauts killed. I don't believe we will surpass that grim record until Starship starts carrying humans.
By the time Starship carries humans (assuming that it will, which I do), it will likely have flown uncrewed more times than Shuttle + all prior American programs together.
They know it’s going to be tricky to make landings safe for crew. Nobody’s in a rush to fly people.
Agreed! But just as airliners eventually had a disaster, Starship will eventually have one too. Maybe it will be earlier and kill only single-digit number of astronauts. Or maybe it will be later, when it is transporting 100 people at a time.
The only way Starship won't kill people is if it's phased out earlier than expected (earlier than I expect, anyway).
And, of course, I very much hope I'm wrong and no one ever dies on a rocket again.
STS-27 didn't involve an abort but it's worth reading about as well due to its direct relationship to the Columbia disaster: https://en.wikipedia.org/wiki/STS-27
Just two missions after Challenger, the orbiter experienced tile damage much like Columbia did 15 years later. It left an entire field of damage all the way down one side of the orbiter. They inspected under the orbiter using the canadarm as much as they could, and found some early indications of tile damage. Mission Control refused to use a high-resolution unencrypted camera to image under the orbiter due to the classified nature of the mission payload, instead using a lower resolution encrypted link, and dismissed the damage as video artifacts.
The crew wasn't happy at all at this assessment and seriously believed they were going to die on reentry. Mike Mullane believed he would see hot aluminum streak up the windshield like raindrops. Commander Hoot Gibson advised the crew there was no point in "dying all tensed up". Even if Mission Control had deemed this to be inevitable, it was not at all clear how quickly or even if NASA could process another orbiter and evacuate them in time.
The shuttle's wings have a combined elevator-aileron called an "elevon"; when both are in the same direction, they focus as an elevator (controlling pitch), and if they are in opposite directions, they function as an aileron (controlling roll). On reentry, Gibson kept his eyes glued to an elevon gauge to see if aerodynamic forces from a disintegrating airframe were forcing the onboard computer to roll the shuttle, creating a difference in degrees across the elevons. If they had deviated more than a few degrees, he planned on cussing out NASA before he and the rest of the crew died.
Fortunately, the shuttle made it back safely, but the damage was far worse than anyone anticipated. The foam strikes had spared the RCC leading edge panels that would eventually doom Columbia, however, nearly one half of the bottom of the shuttle had some degree of tile damage, and one had gone missing completely in a critical section. That section, by the grace of god, had a heavier plate welded to the airframe that was used as part of a communications antenna. One tile missing anywhere else, and the shuttle would have been destroyed.
Neither STS disaster came out of the blue. Each had earlier clues that were ignored, and that's what both accident reports highlighted.
I think STS had so many high-criticality problems that there was probably no way to salvage it as a vehicle. The only sad thing is that they didn't sunset it after Challenger.
My "favorite" Space Shuttle abort mode is the Return to Launch Site (RTLS) abort[0] because of the complex energy management procedure necessary to pull it off.
As I understand it the energy management of an RTLS is tenuous because, ultimately, the orbiter is an unpowered glider (with a glide ratio similar to a brick). The limits of stress on the orbiter's airframe are a factor, too.
The orbiter's mass, energy (altitude), and downrange position all have to be in a reasonably tight envelope to make the landing without coming up short or overshooting (there's only so much energy you can bleed off doing S turns).
John Young[1] described an RTLS abort working thru "continuous miracles, interspersed with acts of God".
Somebody made a fun (if you like this kind of stuff) simulation of an RTLS abort[2] in Kerbal Space Program.
[0]: https://en.wikipedia.org/wiki/Space_Shuttle_abort_modes#Retu...
[1]: https://en.wikipedia.org/wiki/John_Young_(astronaut)
[2]: https://www.youtube.com/watch?v=Vx94b9nHTHg
NASA proposed STS-1 be an RTLS. John Young wasn't too happy at the idea and compared it to playing Russian roulette.
Also fun fact, STS-1 the Orbiter was found to be so damaged, Young said if they knew about it, they would have ejected
This mission also illustrates how confusing the old STS mission numbering scheme used to be.
STS-51-F is the only exercised launch abort, and the Challenger was the orbital vehicle.
STS-51-L, which occurred the following year, is the famous Challenger disaster.
Originally, the numeral in the mission designation indicated the federal fiscal year of the mission and the launch site (51 being 1985 and launch site 1 (KSC, 2 would have been used for Vandenberg). The letter indicated the sequential launch index, so A=1m, B=2, etc.
This was originally done after STS-12. Superstition around Apollo 13 created the desire to avoid an STS-13. The numbering system returned to a strictly sequential schema after STS-51-L, with STS-26.
> Booster Systems Engineer Jenny M. Howard acted quickly to recommend that the crew inhibit any further automatic RS-25 shutdowns based on readings from the remaining sensors,[6] preventing the potential shutdown of a second engine and a possible abort mode that may have resulted in the loss of crew and vehicle (LOCV)
Incredible that human action during ascent (when everything is usually automated) potentially saved a crew. One of probably dozens of cases, is my guess.
STS still has the record for most astronauts killed. I don't believe we will surpass that grim record until Starship starts carrying humans.
By the time Starship carries humans (assuming that it will, which I do), it will likely have flown uncrewed more times than Shuttle + all prior American programs together.
They know it’s going to be tricky to make landings safe for crew. Nobody’s in a rush to fly people.
Agreed! But just as airliners eventually had a disaster, Starship will eventually have one too. Maybe it will be earlier and kill only single-digit number of astronauts. Or maybe it will be later, when it is transporting 100 people at a time.
The only way Starship won't kill people is if it's phased out earlier than expected (earlier than I expect, anyway).
And, of course, I very much hope I'm wrong and no one ever dies on a rocket again.