TFK I think your two NIST quotes are new, at least to me, and very interesting. These minor verbal errors happen all the time in technical writing.
I'm still hoping you or someone can clarify my questions in my post here. These are some interesting ones for me, at least. But I'll admit that the question about whether moment frames are 100% rigid and inflexible is kind of, uh, rhetorical.
I'll defer to Architect on the question of moment frames when used in buildings. Buildings have a lot of consideration beyond just plain strength in their design.
[BTW, terminology. In this context, a "moment" is a "twisting or torquing force".]
But in other types of structures, moment frames are pretty weak.
A moment frame is a rectangle.
The advantage of a rectangle in a building is that you can put windows into it & see out, and it costs a lot less than a truss frame (see below).
The failure mode is when the rectangular frame turns into a parallelogram, by the top member sliding to one side & the two vertical members rotating to an angle. (This is the way that a bookcase would collapse if you removed the back thin sheet of wood, particle board or pressed cardboard, whose job is to prevent this rotation.)
The resisting moments necessary to prevent the rectangle from collapsing to the side must be provided by short extensions (i.e., reinforcements) at the connections. But they have such short "moment arms", that the assembly generally turns out to be comparatively weak. If you used pin joints to hold together a moment frame, the structure would be unstable & collapse. ETA:
Just like one of those expanding wooden gates that we used to put at the tops of stairs to keep young kids from falling down them. (Rectangles, with pin joints. bad example. -tk. Better example: free collapsing & expanding parallelograms that navigators used to use to transfer heading from compass rose to flight path. )
The point is that the strength of a moment frame structure depends on the strength of the moment resisting connections at each intersection of beam & column.
In contrast, a much stronger frame member is a truss, which is a triangle. Or rather, a collection of triangles. In a triangular frame, all 3 end points are completely constrained, and are immobile unless one of the members fails. This is exactly why you see all cranes built with triangular truss members in the extendable arms that have to carry the weights.
Equally important, in a truss, you don't depend on moments being generated at the vertices of the triangle to provide the strength of the frame. One could use (free rotating, zero resisting moment) pins at each vertex, and in principle, the frame would be just as strong as a welded assembly. (In reality, it'd have a lot more "slop" in the structure due to clearances, but the strength should be approximately the same as if one welded the members.)
One problem with truss supports is that you can't see out the windows without having a diagonal member break up your view. And you need to use a lot more material as well, which increases the cost.
If you look at fig 2-25 of NCSTAR1-9 (pg 31, pdf pg 75), you'll see that they added diagonal members to the rectangular moment frames in order to form the two "belt trusses" all the way around the building at the bottom floors & midway up the building in order to stiffen & strengthen the structure. Take a rectangular moment frame, and turn it into two trusses by adding a single diagonal member, and the frame will be much, much stronger.
Just a caveat here, you should solicit Architect's take on this too.
With regards to your other questions:
ChrisMohr said:
did NIST measure the rate of collapse of the north perimeter roofline from the better-known straight on view of the north face? Or did they use THIS video, kind of from the side?
NIST used the same one that Chandler did: the straight on view (aka "Dan Rather" viewpoint.)
They did NOT analyze the descent of "the roofline". They analyzed the descent of one point on the roofline: the north west corner.
ChrisMohr said:
Because if they used the famous north-face-straight-on view, would the "at freefall" measurement have been an illusion as the building was also falling backwards?
Nope. Not due to falling backwards.
But it is important that you understand something: Neither the northwest roof line point, nor the external frame of the building, fell at a constant acceleration.
By choosing to do a "linear interpolation" of the velocity data, both Chandler & NIST FORCED the results to produce a (false) "constant acceleration". Their choice of a linear interpolation automatically hides any details of varying acceleration.
However, if you look at the raw data, you can immediately see where certain adjacent data points have a slope (if you connect the two adjacent dots) that is greater than the average (implying a higher acceleration than "g"), and other adjacent data points that have a slope less than the average "g".
The finer detail analysis that I & others have done show a much closer approximation to instantaneous acceleration by analyzing every frame rather than every 7th or 8th frame (NIST did 4 data points/sec, 1 out of every 7.5 frames, or 1 out of every 15th field).
When you analyze all the frames, it becomes obvious that the descent of that point was NOT any constant acceleration, much less "g". It varied below & above "g". In addition, the acceleration gradually increased, rather than jumping immediately to its absolute value, as something that was really in free fall would do.
ChrisMohr said:
is Chris7 right when he says that moment frames helped hold at least the three visible walls of the building together during its collapse (I'm guessing yes)?
Bad guess.
No, he is not right. He just wishes it to be so. He's not provided any data to show this.
NIST explicitly states that THEIR MODEL (not reality, their model) shows something entirely different: that the west face exterior columns buckled first (near column 14, the southern most column on the west side of the building).
"... exterior columns adjacent to the seven columns severed in the southwest region due to the collapse of WTC 1 were the first to buckle because additional load was distributed to them following the debris impact damage."
NCSTAR1-9, Vol 2 pg 586 (pdf pg. 248)
Then they say: "The south and west exterior columns buckled first, followed by the north and east face columns."
This is mutually exclusive of Chris7's claim. In as much as NIST are a bunch of structural engineers, and Chris7 is not, I'll go with NIST.
I don't know of anyone who has done an analysis on all 3 sides of the building, comparing the timing of multiple points on the roofline to the descent of the northwest corner.
The person that may have done so is femr2. He doesn't seem to post here anymore, and based on history, I'd be leery of taking any analysis of his at face value.
I believe that he is honest, but I am also certain that he frequently ignores significant effects, and greatly overestimates the precision of his measurements.
ChrisMohr said:
What about his claim that these moment frames hold everything together rigidly in 100% alignment with no part moving even a centimeter up down left or right or in or out as long as the moment frames are there (I'm guessing no way Jose)?
Much better guess.
Again, this is merely his wishful thinking. The building was disassembling. There was no possibility that the strength of the moment frames could resist that disassembly, as proven by the fractured components lying on the ground after the fact.
All of this will become moot, and yet will hold together logically, when I explain why the one NIST quote, "the
entire building above the buckled-column region moved downward as a single unit" has to be wrong.
And their quote "the
remaining exterior structure above began to fall vertically as a single unit" has to be right.
Tom