... (putting aside that Tony never actually shares those calculations for scrutiny)."
Quoted for Truth.
Case in point:
It was over at metabunk, where Tony, I and some others (mostly or solely "debunkers", I think) discussed an aspect of the Nordenson study - Nordenson claimed that the potential energy of a floor section, that's falling after the infamous c79-c44 girder walks off its seat at column 79, is sufficient to first deform, break, bend conrete floors and steel elements and then punch out the c79-c44 girder one floor below, to send the floors into a cascading collapse. Tony argued that Nordenson missed one big energy sink, namely that the entire descending floor frame would bend on impact, and that would eat up a lot of energy. The all-important parameter was the "stiffness" of that floor frame - Nordenson treated it as essentially infinite, which results in no energy dissipation through bending, when in fact it has to be finite, and must eat energy. But how much? We developed the formulas, put in some reasonably well established numbers for mass etc, and the remaining free variable was, again, this "stiffness", which can in principle be computed from material properties of steel and the geometry of the frame, but doing this analytically is infeasible. Tony convinced me that modelling this frame with FEA software can get the result - FEA can find out the frequency with which this assembly would "swing", which is directly related to the stiffness. So he did plug stuff into his FEA software, and out came a stiffness value that gave us a conclusion "collapse progression highly unlikely".
I asked Tony to show his model - the input data, or perhaps explain in words exactly what he modeled. Boundary conditions, element types, yada yada. Not like
I could have scrutinized any of this, I have not the slightest idea how to vet a FEA model. I had hoped to get some other engineer to do the scrutinizing for me.
There was no risk for Tony - the other engineers present in that discussion didn't seem inclined, or able, to do it.
And yet, Tony never revealed what he did. Essentially, he through secret ingredients into a secret black box, and got the result he needed and wanted. D'uh.
Don't get me wrong - I really trust that his model was okay, at least as a draft version, I don't allege he fudged things to get what he wanted.
So there is an example where Tony actually scored a win - he convinced me! - and yet he refused to share calculations for scrutiny. Something he and AE never tire accusing NIST of.