Comment by pdonis

> What would happen "normally" (i.e. the normalization of deviance) was that the rotation (from the SRB joints bowing--essentially "ballooning") would create a gap, and the O-rings would get blown into that gap and ultimately seal in there

But data from previous Shuttle flights showed that even that wasn't happening, at temperatures up to 75 F. And the Thiokol engineers had test stand data showing that it wasn't happening even at temperatures up to 100 F. In short, that joint design was unacceptably risky at any temperature.

It is probably true that the design was somewhat more unacceptably risky at 29 F. But that was a relatively minor point. The reason the cold temperature was focused on by the Thiokol engineers (who were overruled by their own managers in the end, as well as NASA managers) in the call the night before the launch was not that they had a good case for increased risk at cold temperature; it was that the cold temperature argument was the only thing they had to fight with--because NASA had already refused to listen to their much better arguments the previous summer for stopping all Shuttle flights until the joint design could be fixed.