Travis,
I've actually just been thinking about the same thing from a slightly different perspective, and have created a spreadsheet to map the increase in kinetic energy and momentum as each floor was added to the dropping mass.
Does anybody know what each floor of the WTC towers weighed as laden on the day?
Much of that calculation can be done without any actual values in lb or kg, as the mass will cancel out in almost all equations. Just take "1 MassUnit (1 MU)" instead of "x kg".
As for the forces (if forces appear at all in your calculations) that structural members can exert without plastic deformation, you can use terms that include a multiple of 1.5-3 of the static load, as that would be a convenient Demand-to-Capacity ratio.
The most interesting numbers, in my opionin, are those that describe the distance that a member kann flex in elastic response before it buckles and loses its load-carrying capacity. This could perhaps sometimes be expressed as a ratio of static load compression : max load compression, and is certainly a number closer to 1.0 than to 0.9
Anyway, my first 1D-model that I did a while ago went with assumptions like "all the dynamic load goes 100% vertically into the columns", "columns can compress to 90% original lenght, carrying 300% of static load, then they break and go down to 0%", etc.
Note that absolutely no engineering went into this - I modeled columns as entities with dicrete physical properties geared on the one hand towards simplicity of computation, and on the other hand calibrated such that all assumptions fell on the side of overestimated survivability. One such extreme assumption is of course that it's the breaking of
columns which propagates collapse. Obviously, this assumption is false; but replacing it with assumptions closer to reality (that mostly see floor-to-column connectors failing) makes survivability less like and acceleration / speed higher.
Crude model, didn't take into consideration transfer of momentum, but still I came suprisingly close to the "real" net acceleration off collapse speed.
Later I modeled the same, this time ONLY considering transfer of momentum (is much easier to calculate) - and came to results that were almost the same - a little slower, IIRC.
Which led me to suspect that the mechanism behind the slowing down of collapse through mere momentum transfer is the same as the mechanism that destroys the structural supports, and of course it is: Momentum transfer happens largely through inelastic collisions, in which kinetic energy is converted to heat (a little bit) plastic deformation and fracture (those two dominate). Since mere momentum transfer predicts a smaller acceleration than mere consideration of structural force, I think that buckling of supports is mostly already included in the energy transfer by inelastic collision, and doesn't add to the resisting forces already included in the momentum transfer.
It follows that modelling the pancking collapse as a series of (110-15) inelastic collisions of top block / rubble with intact floors gives you estimates of energy dissipation and collapse time than mere consideration of structural forces.
Anyway: The answer to Travis is: There is a maximum momentum and kinetic energy of the impacting (top) part that the impacted (bottom) part of intact structure can absorb until it gives way completely. My simple model, which overestimated elastic forces, showed that, IF the top 15 floors were allowed to drop through the height of 1 floor, then the next floor would be overwhelmed by an order of magnititude (factor 10) - which in turn means that at > about 10% of the kinetic energy would already be sufficient to kill the next floor. After that, we will see the equivalent of nearly full floors of unresisted fall, and the overwhelm factor will quickly grow.
Total collapse becomes inevitable.
I am happy to find that Professor Bazanz agrees with me
The problem to solve is just: How did collapse initiate in the first place such that the top 15 floors picked up a velocity > sqr(10) larger than what is attained in free fall through one story, while the intact structure would NOT YET be overwhelmed. (The key to the answer lies in the fact that the respective structure wasn't intact after plane crashes and fires

)