If WTC 7 was a natural collapse (gravity acting on debris and fire damage) vs. an artificially induced controlled demolition collapse, why did it maintain such a relatively level roofline?
Welcome back. This is actually an interesting question. You've received a couple of terse replies that I basically agree with, but I will elaborate. To those bickering in my thread, cut it out, guys. Thanks.
If failure was due to a critical column failure originating in a well off-center location, would not the logical expectation be for that initial failure, to at the very least, lead the collapse?
If you agree with that, would this portion of WTC 7, which is leading the collapse not pull the other intact side down with it---in a delayed, unbalanced manner?
Would that unbalanced failure not have produced a dramatically non-level roofline, as well as a probable topple with the likely possibility of leaving a portion of the intact structure still standing?
To answer this question, we need to make some assumptions about both the initiating failure event and the condition of the structure at that time. Both of these are going to be somewhat speculative. The NIST Report, which as you probably know is supposed to be out by the end of Summer, may provide a significantly different hypothesis.
We know a few things about the collapse. One is that the interior almost certainly failed first, and we know this because the collapse of roof structures preceded motion of the exterior wall. Another is that the collapse initiated low in the structure, which is why a large portion of the perimeter all moved downward as a unit, rather than the progressive mechanism seen in the Towers.
The core of WTC 7 was not entirely different from the Towers -- it was dominated by a small number of massive columns, with connections to the exterior wall spanning an unusually long distance. One of the problems with this design, as noted by Mr. Scheuerman whom I've discussed this with in some detail, is that the core is relatively susceptible to single-point failure. In normal operation, there is no reason why any column should fail, and this is why such a design is acceptable in ordinary practice. But should a column fail for whatever reason, the remainder of the core has very little ability to resist side loads. If the other columns cannot support the weight and the sudden eccentric load, the first column failure can literally pull the rest of the core sideways and then down along with it.
In the WTC Towers, one of our key pieces of evidence for the collapse mechanism is the inward buckling seen at various parts of the perimeter. This indicated severe eccentric loading appled to the perimeter by buckling floor trusses. WTC 7, as far as I know, did not demonstrate this behavior, but instead showed us something different -- a three-story "bulge" in the perimeter, reported by firefighters on the scene well before collapse. This "bulge" is indicative of a similar but slightly different mechanism. Instead of the floors drawing the perimeter inward, the building as a whole is sagging, with the "bulge" marking the shortened side of the structure. Rather than just pull in the perimeter at a point, what we have here is the core being pulled towards the perimeter, and where there is still some stretch opposite this pull a bulge appears.
So with the setup out of the way, let me answer your questions. At the moment stability is lost, the core buckles first -- a column gives way, and the neighboring columns cannot handle the sudden increase in their own load. This failure occurs well down in the structure, and momentum starts to accumulate.
This load next passes to the perimeter columns via whatever parts of the floor survive -- downward motion in the core above now acts in tension on the perimeter, in a fairly even manner. The floors are bent downward at their attachments. So now we have a race condition. Will the floors give way first, leaving a shell of perimeter columns while the core tumbles completely? Or will the floors remain attached, leaving the perimeter columns to fail at their weakest points?
The latter seems to be what happened. In addition to the roof giving way first, there were also "ripples" observed in the perimeter just before total collapse, indicating flexure in the structure. This might be due to "snapthrough" as the core load was suddenly relieved, but I think it is more likely due to an
increase in load than a sudden decrease.
An increased load on the perimeter columns will cause them to buckle, and this will almost certainly happen at the same low level as the initial core failure. There are two reasons for this. First, the perimeter was already deflected here, as the bulge demonstrates. The "kink" low in the structure is the maximum stress point in the perimeter. Second, where the core fails is also where the floor support is weakest, so the perimeter is least braced in this location.
As a result, what we expect is an "upper block" of perhaps 30 stories or more, with the core leading the collapse, detached from the lower floors. Many of the perimeter column failures will be along the same line, low in the structure. Probably not all of them, but most, and the few that remain will still be bound to the other perimeter columns by spandrels. Thus, we do not expect the different perimeter walls to fall at different times, but instead the upper perimeter wall should fall more or less as a unit.
It is, however, still possible for the upper portion to
tilt. And it did. But the situation is much different here than in, say, the WTC Towers. In that case, both upper blocks rotated by a noticeable amount as they started to fall. In the case of WTC 7, the structure also rotated, but not nearly as much in terms of angle.
This is explained through simple geometry. Both WTC Towers upper blocks were comparable in terms of width and height. In WTC 7, the upper block is considerably taller than it is wide, particularly along the short axis. A small angular rotation requires a proportionally much higher unbalanced force in the perimeter wall. This simply isn't expected. There is no expectation that one of the perimeter walls would survive, essentially supporting the entire upper mass of the structure, while the core and other three walls had failed. There is some, of course, and this is why the facade wound up laying on top of the debris pile, but we only expect a degree or two in this case rather than ~8 degrees, as in the Towers.
Now, as you remark, we do expect the off-center initial failure to lead the collapse, but only just. The collapse of the roof before the perimeter is evidence of the core failing first. The "kink" in the upper roofline is evidence of locally higher stresses at one point of the perimeter, and evidence of that stress being counteracted by spandrels and floor structures pulling down neighboring perimeter columns. The slight lean and rotation of the structure as it fell is evidence of an off-center collapse initiation, but the structure is simply too tall relative to its footprint to support a large amount of rotation -- the "hinge" will fail very quickly, before it can impart much angular momentum.
So the answers to your questions are "yes, but only slightly." None of these features is unexpected in an unassisted building collapse, due to its design. The above would not necessarily be true of a more conventional post-and-beam structure, a nested-tube design, or a masonry structure.
Regarding some remnant remaining, since the collapse initiation appeared to be low in the structure, any remnant would have to survive being hit by the falling mass of the structure. This would be sheer luck. Above the collapse initiation, neither the core nor the perimeter, nor any fraction thereof, would be likely to survive this kind of collapse. The two systems are braced by each other, and singly are simply not stable.
I apologize for the degree of speculation in this post. I look forward to the NIST Report in helping refine the possible space of explanations.
