arcniz, someone at the top is finally waking up to reality, just like they did after Challenger. Then, they realised that the ELV program was almost non-existent and took steps to rectify it. Now, they find that only three Orbiters remain to look after the ISS/HST and other great investments. Aviation is not isolated from the rest of history in repeating itself every so often. The politicians see to that!
OVERTALK's sacrificial strip is an interesting concept. I feel, however, that preventing bits falling off the ET during launch might have first priority, as the TPS has worked satisfactorily on 111 occasions and has been upgraded in the light of operational experience.
On the question of ET ice, I remain doubtful. It was not a cold day and those who were around the stack prior to launch say that there was no visible ice present on the tank. Also,
OT, the ET coating is weatherproof and,
theoretically cannot soak up water. I understand that there were no signs of cracked foam on launch day. However, the thermal cycles you mention are valid. What if
micro-cracks evolved that were not easily visible but let in some moisture?
Ice has always been an issue on the ET, but
large chunks of foam falling off has not, until STS-112/107. There may be other issues here, perhaps the earlier redesign of the foam to eliminate CFCs, for example. I hate to raise it, but QC has to be considered too. As I said on an earlier post, without the ice, one 2.5lb mass (or two!) striking at just the wrong angle, at that speed, is quite a whack.
With all the descriptives flying around about the RCC sections (my favoured theory) I thought it might help to post a diagram which illustrates the RCC attachment system. This is from the Curry/Johnson
presentation (pdf 900kb) at the Future Space Shuttle Conference 1999. The enlargement also shows the later mod and a T-seal. To my mind, the (flexible) T-seals are one feature that would not enjoy being whacked and could let in a lot of plasma, initially in a metered manner.
Enlargement
As mentioned elsewhere, if plasma gets past, or through, an RCC panel, how long would it take to reach past the secondary protection and on to the alluminium attach points? Where it goes thence is largely academic - but surely a "bad day."