The space shuttle, like most launch vehicles, was a multistage vehicle (albeit one where the various stages were mostly stuck to the sides of each other rather than being stacked vertically), reducing the amount of mass and weight needing to be carried all the way into orbit:
- The first stage was the two SRBs, which burned from liftoff until fuel exhaustion about two minutes later, at which point they fell off the sides of the shuttle stack, were pushed away by sepratrons at their head and tail ends, and parachuted into the ocean for recovery and reuse.
- The second stage was the three ET-fed RS-25 SSMEs on the orbiter’s rump, which burned from several seconds prior to liftoff until shutting down approximately eight and a half minutes after liftoff; following engine shutdown, the ET fell off the orbiter’s belly, reentered the earth’s atmosphere, and despawned over the South Pacific.
- The third stage was the two AJ10s in the OMS pods on the orbiter’s upper rump, which burned twice: first for about one-and-three-quarters minutes, starting a few seconds after SRB separation, and then again for between thirty seconds and a minute, starting about eighteen to twenty minutes after liftoff, to perform the final orbital insertion. (In the event of a severe underspeed, another burn could be inserted, this one up to several minutes long, starting ten to eleven minutes after liftoff.)
Although the SRBs and ET separated properly every time they were required to do so throughout the duration of the shuttle fleet’s entire service life, failures-to-stage (both with side-mounted strap-ons and with vertically-stacked core stages) have occurred many times with expendable launchers,1 and there is no reason to assume that they couldn’t have happened with the shuttles; yet, the space shuttle flightcrew operations manual contains no information on what to expect or what actions to take in the event of an SRB or ET separation failure, not even in its otherwise-extremely-comprehensive “Emergency Procedures” section.
Given that the shuttle stack would already have been very high in the atmosphere by the time of SRB burnout, and would have been well and truly in space by the time ET jettison was commanded, the aerodynamic forces on the vehicle stack would have been low to negligible even with a retained SRB(s) or ET, and, thus, I’m assuming that an SRB or ET separation failure would not have led to an immediate loss of control. The main effect of a separation failure would, presumably, have been the weight and mass penalty of the non-separated components:
- In the event of an SRB separation failure, the added deadweight would, most likely, render the shuttle totally incapable of reaching orbit, probably requiring either an RTLS or TAL abort (which one would probably depend on the mission’s trajectory, gross weight, and Δv margin); a TAL would probably be preferable, if at all possible, because a) a TAL was a much simpler and more forgiving abort mode than an RTLS, keeping the vehicle much further away from the edges of its flight envelope than an RTLS would, and b) the aerodynamics during the late stages of RTLS powered flight (with the vehicle at relatively low altitude, moving back towards the launch site, and powering out of a momentarily-vertical freefall), and during the extended mated-coast phase used for an RTLS, might be rather iffy with one or both SRBs still attached to the ET, potentially risking a loss of longitudinal or directional control or a collision with the jettisoned ET-cum-SRB(s). In either case, the ET (with SRB[s] still attached) would presumably be jettisoned at the normal time for whatever abort mode was chosen.
- In the event of an ET separation failure, the vehicle - already quite close to full orbital speed - might (depending on the mass of payload carried) have been capable of reaching at least a low ATO orbit with a set of extended OMS burns (possibly using the orbiter’s aft RCS jets to assist); this answer, concerning earlier plans to bring an ET all the way into orbit for use as a space station or whatever, indicates that a shuttle could potentially have reached orbit without jettisoning the ET, at least under the right conditions (light payload, no other performance-reducing occurrences during launch).2
What were the procedures, if any, for responding to an SRB or ET separation failure?
1: Including on at least one manned mission.
2: Pressing uphill to some kind of orbit (even a low one), and then either waiting for rescue or attempting an emergency EVA to manually free the orbiter from the ET, would pretty much have been the only even-potentially-survivable option for an ET separation failure, as reentering with the ET still attached would have (not necessarily in this order):3
- Led to a loss of control, due to the aerodynamic forces on the ET.
- Caused the ET to explode from shock heating (via the boiling and thermal expansion of the residual propellants in the ET), likely damaging the orbiter (especially its lower surface, home to important parts of the orbiter’s thermal-protection system) in the process.
- Likely blocked the orbiter’s landing gear from being lowered.
All of these would probably have been bad.
3 As a result, if (say) the ET were to fail to separate during a TAL, the astronauts would basically be screwed.