Originally posted by ANOK
Originally posted by IrishWristwatch
How is that showing the FoS of components, and the combination of components?
It's an idealization, of course. Whether or not this approximation is valid is another matter. The claim was made - multiple times - that Bazant
ignored factor of safety. He did not.
That shows columns of one story, the columns were not separated by floors, they were continuous from top to bottom...
Again, it is a simplification; now, let's see how good of one it is. The perimeter columns were pinned top and bottom at every story by the spandrels
and floor assemblies, likewise the core had horizontal cross connections and was bounded by the floor spans at each level. So, yes, I'd say it is a
good approximation (for the mechanics) in that, despite the three story unit column lengths and essentially contiguous assembly path all the way down
the structure, the columns indeed have effective length of one story.
Your alternative is to ignore the lateral bracing and pinning, either by hand wave or by virtue of connections being destroyed during the collapse,
but that acts to make the same columns weaker
, not stronger. Slenderness ratio.
But the fact is, buckling (either perimeter or core) was not observed on a large scale, so the question is moot. The dominant failure mode appears to
be connection failure. Much of the perimeter and core simply fell over in great unsupported lengths, immediately obvious in many videos. If the mode
had been buckling, it would've required much more energy, and Bazant's scenario of buckling over one-story lengths maximizes
dissipation. Seemingly the opposite of what you think should be the case.
...and tapered in size significantly which means the FoS changes considerably...
As I said, it doesn't matter how strong things are "down there" if "up here" the columns are pinned every story and are constrained to buckle (again,
if they were to actually buckle) in story lengths or less. All the same, stop and think for a second: FOS does not necessarily change considerably
going down. A constant FOS means capacity scales with imposed load, so simply having matching capacity in members does not increase the FOS.
Ideally, FOS would the be the same all the way down regardless of the heterogeneity of mass distribution. An FOS of 2 would support twice the static
load no matter the magnitude of load, no matter the location in the structure.
What is really important is the Fos of the combined floor truss connections, because that is what the OS claims failed.
Well, I'm not going to get into defending the grander OS. All I wanted to do was point out some false claims regarding Bazant. I don't even care too
much for Bazant's work as a description of the collapses, but a spade is a spade. The engineering work in there is excellent. The political
statements in an engineering journal? Uncalled for. The applicabilty to the collapses? Eh.
What pressure the connections could withstand before failure is what is needed to be known.
If I understand what you're saying correctly, I agree.
Bazant considered none of that.
Bazant made mention of fracture as the likely dominant mode of failure. He chose to ignore that path because it categorically results in lower energy
dissipation than hinge buckling. Since one of his objectives was to provide a bounding case, this acts in support of that objective. It would be
more accurate to model a collapse using connection failure as mode, but a hell of a lot harder. If there's no expectation that it increases chance of
survival, the work would not act to define the boundary case.
edit on 28-10-2011 by IrishWristwatch because: (no reason
edit on 28-10-2011 by IrishWristwatch because: clarification of point about FOS; typos
edit on 28-10-2011 by
IrishWristwatch because: more typos