The Great Thermite Debate...

[eomer]

115. You make some good points, let me try to construct my position better.

Agreed that it's not impossible to test parts of these models against the real world. But I still say that it was impractical to test them sufficiently.

In many engineering projects the models will eventually be tested "fully" do to the fact that the goal is to design and build something. By "fully" I mean that in addition to testing parts of the model along the way, they will ultimately test the whole system after it has been built. But even when tested "fully", the system may still not be tested fully.

What I mean by that is that even when they have tested the entire system multiple times, they are still sometimes wrong because their ability to test very complex things suffers from similar limits as their ability to design very complex things, which I think is related to the fact that, as you say, they are always, by necessity, using simplifications.

As an example, I'm sure that the o-rings used in the space shuttle program were extensively modeled and tested as components and the entire shuttle launch system was tested multiple times in real use before a disastrous accident made it clear that the design had flaws. The tests of the system had never included testing under the specific conditions (temperature, etc) that made the o-rings fail. Things that they didn't know made what they thought they knew from modeling and testing wrong in ways that were fatal.

And I think this is a pretty good analogy for what I'm trying to say. If the engineers designing the space shuttle launch system were not able to develop a sufficiently reliable model for a relatively simple component like an o-ring, even though they surely tested at multiple levels and stages, then I don't believe that engineers can reliably design and test a model for the WTC tower disasters, which seems to me to be many degrees of magnitude more complex.

And besides being more complex than an o-ring, the WTC tower model cannot be ultimately tested in its entirety since no one is going to build another one and rerun the disaster. In a more subtle way the space shuttle has the same problem in that even though it was built and tested it was still not possible to test it in every way that would be required to give complete assurance. So besides being much more complex, the WTC tower model can only be tested much less thoroughly than the o-ring and even so the testing of the o-ring wasn't enough to rely on it.

In much fewer words, solving this particular problem is beyond mere mortals.

Edit to add: and I need to circle back to how I still think circular logic is used regarding the WTC models, but don't have time now, will have to come back to it later.