[Frank_Discussion]

...why a mere 60 degree rise on the external fuselage above the left wing was significant.

If enough heat got past the tiles to raise the internal structure 60 degrees in five minutes, the tiles aren't doing their job. Internal heat changes are normally very slight on the structure of the shuttles themselves.

The sensor drop-out is easy to explain, if you understand something of how sensors work. The computers ask the question of a sensor: "What's your reading?" The sensors do not send data, per se, but respond to the question. If a burn-through obliterated the sensor between questions, the next question simply gives no answer.

[mikegi]

If enough heat got past the tiles to raise the internal structure 60 degrees in five minutes, the tiles aren't doing their job. Internal heat changes are normally very slight on the structure of the shuttles themselves.

But they said that the other side registered a 15 degree increase. So, an increase in temp was expected. If a 15 deg increase was expected, then a 60 deg rise would certainly not cause problems. In addition, apparently no other sensors reported a rise in temp. Why not?

I'm well aware of data acquisition, having written software to do it for monitoring industrial systems. You typically get a bunch of samples and average them over a given time period in order to generate a result (tossing outliers, etc.). Do you know what the sampling rate was on the shuttle's temp sensors?

[Ramius]

If enough heat got past the tiles to raise the internal structure 60 degrees in five minutes, the tiles aren't doing their job.

Yesterday they said that it was an external temp reading that went up 60 deg in 5 min. They also said that an internal temp sensor in about the same place had not registered a significant change.

[lepton]

If a burn-through obliterated the sensor between questions,

Or the communications cable/wire.