Another scenario

I’ve now seen more (but still sketchy) details on the telemetry timeline in the port wing. There is enough there for me to suggest another possible scenario.

The first problem detected was a temperature rise in the port hydraulics. All the flight surfaces on the shuttle are controlled by hydraulics. Pressure is supplied by the APU’s; control is supplied by actuators controlled by the shuttle computers (GPC’s). There are 5 GPC’s. If I remember the architecture correctly, each controls a seperate hydraulic loop. If a GPC fails and tries to ram an aileron full down, the other 4 override it. This is not handled in electronics, it is handled in the hydraulics themselves. The pressure from 4 actuators pushed by hydraulics one way over rides the one going the other way. So long as no more than two GPC’s fail unsafe, control is still possible.

Let’s posit a heat induced hydraulic failure in the port wing. The temperature rise is a common mode failure which overrides the redundancy. I do not know if they have any additional failsafe to return and lock the position of the aileron at a neutral position. However even if they did we can see a potential problem. The shuttle was in the second of two banks in an S curve the shuttle follows for bleeding off energy. Just at the time communication was lost. If the shuttle has just been commanded to roll and lost all hydraulics when put under the pressure by the actuators, we have the shuttle going into a roll that will go faster and faster. It seems likely (to me) we’d lose the aileron very quickly, followed by breakup within seconds.

Even on level flight, I could imagine serious problems from complete loss of use of control surfaces on the port side. I doubt the fly by wire system could deal with something that extreme. I doubt you can fly a brick that way. Period.

CORRECTION: After chatting with another knowledgeable friend and doing a bit of checking I found I was in error about the number of loops. There were 4 loops in the Enterprise drop test article; they apparently cut this down to three independant loops for the first flying article.