But that's the result I got - I presented my argument in full, the assumptions, the graphics, the formulas used and the algebra done. If there isn't anything wrong with any of these, then the result is correct.

This appears to me as a good, correct way to model the steel frame of the floor, and I understand crudely that FEA can give you a natural frequency and that natural frequency is leads to stiffness via the formula you provide here:

You lost me here somewhat:

1. I have to think about slugs - damn the USA and her imperial units!

2. "just over 20,000 lbs" is over 21,000 lbs, isn't it? Girder is 547 inches * 133 lbs/12 in = 6063 lbs. Each (?) of the 5 long beams is 644 inches * 55 lbs/12 in = 14758 lbs total. I am at 20,821 lbs, add to that the three support beams connecting to the north wall (ok, that's a nitpick)

3. At some point, we'll have to wedge in the mass of of the floor slabs...

4. K = 6,633 lbs./inch is precisely the same result you got in post #97, albeit with a different frequency and different mass (using Nordenson's mass of 46,000 lbs that includes floor slab). Weird coincidence?

The calculation is correct for K = 6,633 lbs/in.

**Before concluding too much, I want at least three of us to be on the same page.**
My main problem at this point is that I cannot check your FEA - I don't have all your files, and if I had them, I wouldn't have the software to use them; and if I had the software, I'd lack the expertise using it.

**Sooo who is here to help and offer a criticism of Tony's approach?**
Also, I question (

**this is an open question to all**) if it is ok to ignore the floor slab in the FEA. Sure, I ignored the slab, and the beams, in my own model of "rotating girder - 2/3L load point". This is a difficult thing to model, for the slab has already broken and partially disconnected.

**I think what we can agree on at this point** is that Nordenson's Appendix B would need rework and refinement if it were to be used again, for his rather clear result of "will fail" would shift into "may not fail" territory and possibly beyond. This could be a worthwhile question to be put to Dr. Hulsey and his PhD students.

**What I cannot agree on** is your (Tony's) conclusion that your partial FEA "

*proves that the falling girder would not have sheared the seat*". You make a good case showing that it may not, but there are too many unknowns at this point to be sure.