Agreed. But you do need the downward momentum to apply to whole upper block, right? Otherwise you don't release its potential energy. Until recently, I was sure the dynamic load of the moving block (at some minimum speed on impact with the lower section) was essential to the explanation. But the model of collapse you seem to be working with (see also next quote) doesn't seem to need any initial downward momentum, only a (lateral) shift of the (static) weight of the upper section from the columns to the floor connections.
Here you are conflating collapse initiation and collapse progression.
Initiation necessarily starts out without any, or negligible downward momentum: Collapse initiates when the load bearing capacity goes below
static load and no more path for successful load redistribution is available. From that moment on, the top block accelerates down. This obviously changes the geometry of the structure, and what comes next depends critically on which failures occurred and accumulated where and which parts of the structure are moving towards what direction. It may happen that a structure that has just started to accelerate downward soon runs into a configuration that manages to arrest collapse. Or it may not. This would require capacities at least twice as large as static load, as static load is now applied
dynamically.
At some point (and it comes soon), momentum (or rather kinetic energy) has grown so large, that at no level the structure can absorb and dissipate that kinetic energy plus the potential energy differential associated with the vertical distance over which that energy is dissipated - that's the condition for collapse progression. This minimum kinetic energy or momentum of course is associated with a downward velocity, which in turn can be thought of, conceptually, as the result of "freefall through a height differential dh". dh does not have to be a full story. And it needs not literally be freefall: You can reach the same velocity with half g in twice the time, for example.
Anyway, what you need to get the "upper block" moving initially, all you need is gravity, and the weakening of the supporting structure to the point where capacity < load. -> Collapse initiation.
Once it moves, you are dealing with dynamically applied loads, which multiplies forces -> Collapse progression.
You may argue that immediately after initiation, while there is conceptually a chance for arrest, you have a third phase, which we might call "collapse transition". Not sure if that helps.
Here's my amateur interpretation of what you're saying. (Which the book I'm hoping for would get right in whatever way I'm getting wrong.)
Let's imagine that all the columns are cut cleanly and horizontally at floor level (say, floor 80) and that the footprint of the upper section is somehow shifted about 1 meter north-west (i.e., along the diagonal) so that all the columns (above and below the cut) are offset from each other. That would of course be catastrophic. The perimeter columns of the north and west face would now be supported by nothing at all, most of the core columns would be suspended over elevator shafts, and the south and east faces would be resting (momentarily) on the floor of the 80th storey. That floor would immediately collapse.
Exactly - and you would not even need to shift by that far. even if you shift by only half the thickness of the steel plates that build up the columns, that would reduce their capacity by a factor near 2, and be catastrophic.
After falling the height of one floor, the moving mass would impact the 79th floor. Or, at least, that's what would happen along the south and east faces. On the north and west sides, the ceiling of the 80th storey (i.e., the underside of the 81st storey) would come down on the top of the columns, where the 80th floor is still attached (since there is no weight on those connections yet). Here the weak spot is also the floor connections, but not those of the 80th floor. Rather, it's the 81st floor that would be destroyed, upwards, by the strength of the undamaged columns. While it would look more or less symmetrical from the outside, the destruction is passing asymmetrically through the building internally. To the SE the floors of the lower section are collapsing as the columns of the top section press down. But the NW the floors of the upper section are pancaking onto the tops of the columns of the lower section.
Yep, good mental model.
I'm sure the result is the same as what we saw. And I'm sure I'm oversimplifying it -- in reality it would be much more chaotic.
Of course.
But it's basically that process of floor connections being broken and columns being impacted (probably in some cases buckling) that I'd like to see described by someone more qualified than me.
Ok.
Conspiracy theorists following along no doubt imagine that a lot of the mass of the falling block dissipates as it disintegrates and falls off to the side.
It would be their job to explain how that much mass moves laterally from within the footprint to outside the footprint within the short time that a single floor collapses.
So it would be good to explain how sufficient mass remains within the footprint of the tower to keep the process going. (The collapse front of course gathers mass from the lower sections it destroys. And also some momentum as the speed increases.)
Things fall mostly straight down, don't they?
With each floor that the top block moves down, a floor slab is added to a growing layer of compacted concrete, steel and office contents. Some perimeter panels, core columns and core beams also get mixed into that growing debris layer.
Some perimeter panels get cut off and fall freely outside the lower part of the tower - this can be seen in many collapse videos, large pieces of steel falling ahead of the collapse front. Those are obviously lost to the growing debris layer - but a high percentage of the mass remains inside the footprint and gets heavier and heavier, such that the collapse gets driven by that layer punching out the floor slabs it falls in in sequence.
It can be computed fairly easily, an Excel spreadsheet would do, how such a collapse accelerate if we picture it simply as just floor slabs falling on floor slabs, just considering Conservation of Momentum and Energy, plus gravity, but ignoring structural resistance: It turns out that after just a few floors down, acceleration closes in on 2/3 of g. Which is what David Chandler measured.
So does this mean there was no structural resistance?
No.
When mere Conservation of Momentum dictates that fall rate is diminished from g to 2/3 of g, this means that through the collisions of debris layer with fresh floors, 1/3 of the kinetic energy is dissipated by "destruction": Smashing concrete to pieces, grinding gypsum to dust - and bending and breaking steel. This 1/3 of available energy can be computed - it works out to the equivalent of more than 40 tons of TNT.
This should be more than enough to cause the amount of destruction seen.
If a Conspiracy Theorist wants to argue that the equivalent of 40 tons of TNT is
not sufficient to destroy the towers as seen, then they need to show work and
specify how much more energy they claim is needed. Since it is highly doubtful that they have worked this out to a precision like "
40 tons of TNT is too little, but 45 tons of TNT would suffice, so we need to add only 5 tons of explosives", we'll see quickly that they would have to theorize the use of at least another 40 tons worth of explosives - and then they would have to explain why we don't hear them, how they would be employed, and why no other evidence thereof has been found. You see, in actual explosive demolitions of highrises, you only need "hundreds" of pounds of explosives, in the case of the WTC towers perhaps a couple of tons. But >40 tons?