Merged Discussion of femr's video data analysis

[femr2]

Yes. In femr2's animation, the bowing is to the right. Combined with what femr2 wrote above, I conclude that the right hand side of femr2's animation is supposed to represent the building's interior.

Glad the additional information helped you out. Strikes me as really funny that it would be required within a multi-page discussion about inward bowing

Strange that femr2 wouldn't, because I would say his animation shows buckling to the right, which is inward.

The lower column is springing back outward after failure of the bolt seam between the upper and lower columns. (In direct contrast to pgimeno's understanding of the real-world behaviour being discussed.)

Before anyone draws the seemingly obvious but misleading conclusion that femr2's animation refutes femr2's claims, take note: femr2's animation is nothing more than his drawing of his interpretation based on his data analysis, so it has no evidentiary value.

ROFL. The behaviour of the panels during initiation is contained within the video record. All you have to do is look at it.

On the other hand, the fact that femr2 tried to support his claims by posting an animation that appears to contradict his claims might lead us to conclude something about femr2's argument. I'm just pointing out that the failures of femr2's arguments do not tell us anything about what really happened.

Upon fracture of the WTC2 East wall along the oft repeated staggered path, the lower columns can be seen to spring outwards back towards their normal vertical position, not inwards.

All you have to do is look. (And not forget what you've seen. Y'know, that ol' memory hole )

 

[femr2]

I preferred not to make that assumption when it is easy to ask.

Which is fine, though in a multi-page discussion about inward bowing, I would have thought it would be more than a little obvious.

Why should I assume the direction of the bowing makes sense

Why would you assume it doesn't ? That would be pretty odd and maroon-like.

when the same animation also shows a very rapidly developing (in a small fraction of a second, based on the time scale of the rest of the animation) tilt of over twelve degrees of the entire structure above the break

Oh come on, it's not a realtime simulation. It's a stick figure to get the *spring-back* point across.

I assume you have no problem with NIST writing...
Column splices failed at every third panel and columns sprung back from inward bowing
...except of course that *every third panel* is wrong...every panel.

or, possibly, instead omits rapid lateral motion and another bend of comparable magnitude forming after the break (of which the bend could hardly avoid also breaking and all of which has no apparent cause) out of the frame of the picture above the break?

I'm sure after relatively short period of time, the upper edge (passing behind the lower) impacting with internal structures, gets thoroughly *mashed up*. But what happens *later* is not the point here.

It could well be that IB of the upper columns does not *spring-back* significantly, or perhaps instead increases, though I don't think it's possible to see such behaviour. It only has to descend a couple of feet before encountering barriers after all.

Again, as long as everyone is clear that .... you know

Just trying to assess which aspects of "reality" the animation is reproducing and which it's ignoring, and why. Because, it's not very real overall.

It's clearly marked as a simple animation capturing the basic behaviour of the columns as the bolt seam fails.

The most important behaviour from my perspective is the spring-back of the lower column, not the behaviour posted by pgimeno (who very aptly highlighted why it's still worthwhile bringing these points up, as he clearly hasn't recognised/identified/watched the actual behaviour yet)

Bear in mind that imo initiation is core-led, so the notion of the East perimeter unloading and causing tilt doesn't work. IB ocurred, though likely via a different mechanism than described by NIST. Late rapid increase in IB is caused by tilt imo...

As long as everybody...

Now, perhaps the noise following pgimeno's whoops has died down and I can get back to some video data analysis ?

I'll be doing some traces on other known controlled demolitions.

Any suggestions should include:

a) 720p video links
b) Pretty static camera from as normal a viewpoint as possible
c) Building measurements

 

[DGM]

Which is fine, though in a multi-page discussion about inward bowing, I would have thought it would be more than a little obvious.


Why would you assume it doesn't ? That would be pretty odd and maroon-like.


Oh come on, it's not a realtime simulation. It's a stick figure to get the *spring-back* point across.

I assume you have no problem with NIST writing...
Column splices failed at every third panel and columns sprung back from inward bowing
...except of course that *every third panel* is wrong...every panel.


I'm sure after relatively short period of time, the upper edge (passing behind the lower) impacting with internal structures, gets thoroughly *mashed up*. But what happens *later* is not the point here.

It could well be that IB of the upper columns does not *spring-back* significantly, or perhaps instead increases, though I don't think it's possible to see such behaviour. It only has to descend a couple of feet before encountering barriers after all.

Again, as long as everyone is clear that .... you know


It's clearly marked as a simple animation capturing the basic behaviour of the columns as the bolt seam fails.

The most important behaviour from my perspective is the spring-back of the lower column, not the behaviour posted by pgimeno (who very aptly highlighted why it's still worthwhile bringing these points up, as he clearly hasn't recognised/identified/watched the actual behaviour yet)

So, what you're saying is "it's close enough for you".

 

[W.D.Clinger]

Yes. In femr2's animation, the bowing is to the right. Combined with what femr2 wrote above, I conclude that the right hand side of femr2's animation is supposed to represent the building's interior.

Glad the additional information helped you out. Strikes me as really funny that it would be required within a multi-page discussion about inward bowing

Had you been capable of acknowledging the fact that your animation shows bowing and buckling in the same direction, both to the right, you'd have been more likely to understand our desire for greater clarity.

Strange that femr2 wouldn't, because I would say his animation shows buckling to the right, which is inward.

The lower column is springing back outward after failure of the bolt seam between the upper and lower columns.

So your animation shows inward bowing, inward buckling, and outward springing. Got it.

 

[femr2]

So, what you're saying is "it's close enough for you".

That's not what I'm saying, clearly. What I'm saying is...

It's clearly marked as a simple animation capturing the basic behaviour of the columns as the bolt seam fails.

My preference is to post segments of the actual video record, which...wonders never cease...I did

I guess I was asking too much of some folk to be able to see the behaviour described by actually looking at the actual video record, so treat my stick animation as a visual cue to help guide you to what you're supposed to have already noticed in the actual video record.

 

[femr2]

our desire for greater clarity

Our ? Funny.

I have no problem with delving into greater clarity.

So your animation shows inward bowing, inward buckling, and outward springing.

You think the lower columns are buckling inwards ?

Inward bowing, sure.
A form of buckling as discussed within the thread, sure.

 

[femr2]

our desire for greater clarity

Which raises a few questions (for me)...

What governs the direction of the fracture ?

What governs the direction of an elastic buckle ?

Individual direction for each half of the fractured member ? (which in this case could be the bolts...)


The lower columns clearly spring back.


The cracks in the bolts probably formed somewhere between in-plane sliding (placing the propogation direction parallel to the applied shear stress) and out-of-plane sliding (placing the propogation direction normal to the shear stress). More in-plane I suppose ?

 

[pgimeno]

pgimeno, I think you are off track now.
That popular animation of a column developing 3 hinges shouldn't be taken so far as to imply that any column actually broke in 3 places. They most likely didn't: As soon as the first hinge breaks, strain is released, and the other 2 spring back to normal.

In my animation, the pieces represented complete panels, not panels broken into several pieces. From that point of view, I don't see any contradiction with what is being said here about the breakage happening at the bolts.

In my view, the assemblies were parts of a bigger column, which is the one that developed hinges (although with zero resistance), which is why I used that diagram.

ETA: By the way, in my animation both the top and the bottom spring back.

 

[femr2]

In my animation, the pieces represented complete panels, not panels broken into several pieces.

Still wrong.

Panels don't break off from the facade. Instead the *seam* between the entire upper section and the entire lower section breaks...along the oft repeated staggered pattern.

You can see the top edge of the bottom face break line emerge from the dust here...

http://www.youtube.com/watch?v=xkpOPA45M4c
http://www.youtube.com/watch?v=SpUTXF9isII
http://www.youtube.com/watch?v=nmQRvhYTA3Y
http://www.youtube.com/watch?v=GgnkO1SeuPU
http://www.youtube.com/watch?v=iLkyEvWe9mY

From that point of view, I don't see any contradiction with what is being said here about the breakage happening at the bolts.

You have:

a) bolts between to panels breaking in the middle of your IB region.
b) bolts at the top of the upper panel breaking
c) bolts at the bottom of the lower panel breaking
d) two panels continnuing to *bow inwards*

(c) & (d) did not happen. (b) probably didn't either.

In my view, the assemblies were parts of a bigger column, which is the one that developed hinges (although with zero resistance), which is why I used that diagram.

You have one break that didn't happen, probably two.

My original animation stands.

ETA: By the way, in my animation both the top and the bottom spring back.

Yet your lower panel (which in the real world doesn't break at the bottom, but instead springs back) continues to *bow inward* forming your pet 3-point buckle...which is NOT what happened.

Why is this so difficult for you ?

You can see the edge of the break line in the video record and short clips I'm posting. You can see that the lower edge of the panels haven't broken and are back in line fine and dandy.

 

[Reactor drone]

In your animation, probably not. But this is my interpretation of what is seen...

(Hand-made, in a hurry - excuse the inaccuracies)

[qimg]http://www.formauri.es/personal/pgimeno/xfiles/11-s/buckling-wall.gif[/qimg]

I'd say this was a better illustration of the mode of failure of the connections (tension failure) than Femr2's diagram which seems to show a shear failure of the connections. The only critisism I'd have with it is that it probably has too many complete failure points.

I'd also characterise these failures as being post-initiation. Bowing/buckling is in the initiation phase and breaking of the connections follows as a result of the downward motion of the building post initiation.

 

[femr2]

Femr2's diagram which seems to show a shear failure of the connections.

The lower panels spring back into place...

What breaks in your opinion ? Bolts or column ends ? (There's some words in t'NIST report on the subject, I imagine )

I'd suggest that, depending upon the vertical location of the bolted seam between the upper and lower panels, which obviously there are three *types* of, the exact fracture/shear mode will be slightly different. Possibly 3 for lower face, and 3 for upper.

See YT links above for 720p views.

And, of course, the upper edge of the break line can be clearly seen.

 

[Reactor drone]

The lower panels spring back into place...
[qimg]http://femr2.ucoz.com/_ph/5/2/879008948.gif[/qimg]

What breaks in your opinion ? Bolts or column ends ? (There's some words in t'NIST report on the subject, I imagine )

Bolts would be the weakest points at the ends of the column sections therefore the parts most likely to fail in most cases (rather than weld failure or breaking around the HAZ near the welds)


I'd suggest that, depending upon the vertical location of the bolted seam between the upper and lower panels, which obviously there are three *types* of, the exact fracture/shear mode will be slightly different. Possibly 3 for lower face, and 3 for upper.

Or maybe the same failure mode with slight differences in timing depending on the position of the joints in the buckled wall as the building descends.


See YT links above for 720p views.

And, of course, the upper edge of the break line can be clearly seen.

Answers in bold above.

You seem to be pointing to connection failures as though you're the first person to see them. I think we're all aware that connection failure is the main failure mode of the buildings during the collapses. So the lower part springs outwards following the buckling failure of the wall, that doesn't mean the wall didn't buckle and that inward bowing didn't occur. It basically has no bearing on anything and I'm struggling to see the point that you're trying to vaguely imply here.

 

[GlennB]

Answers in bold above.

You seem to be pointing to connection failures as though you're the first person to see them. I think we're all aware that connection failure is the main failure mode of the buildings during the collapses. So the lower part springs outwards following the buckling failure of the wall, that doesn't mean the wall didn't buckle and that inward bowing didn't occur. It basically has no bearing on anything and I'm struggling to see the point that you're trying to vaguely imply here.

Amen. I was about to write much the same thing.

 

[femr2]

You seem to be pointing to connection failures as though you're the first person to see them.

Nope. Some people are aware of the nature of the failures, some are not. As W.D.Clinger said...our desire for greater clarity.

I've even scoured the NIST reports for any detail on the behaviour, resulting in this one, incorrect, assertion in a summary table...

Column splices failed at every third panel and columns sprung back from inward bowing

Note that it's inaccurate...every third panel ? Nope.

pgimeno still thinks his 3-point buckle diagram fits the behaviour of the panels of WTC2 East face as it splits along the bolt seams.

I think we're all aware that connection failure is the main failure mode of the buildings during the collapses.

Which is fine, but the point is to highlight that what failed was bolts joining two columns together...within a discussion where statements such as "With continuously increased bowing, the entire width of the south wall buckled inward." are being highlighted.

So the lower part springs outwards following the buckling failure of the wall, that doesn't mean the wall didn't buckle and that inward bowing didn't occur.

What governs the direction of the fracture ?

What governs the direction of an elastic buckle ? (around the axis with the lowest moment of inertia I suppose)

Individual direction for each half of the fractured member ? (which in this case could be the bolts...)


The lower columns clearly spring back.


The cracks in the bolts probably formed somewhere between in-plane sliding (placing the propogation direction parallel to the applied shear stress) and out-of-plane sliding (placing the propogation direction normal to the shear stress). More in-plane I suppose ?

Given that the upper column and lower column travel in different directions following failure of the connection between them, what direction would the buckle be in, and why ?

It basically has no bearing on anything and I'm struggling to see the point that you're trying to vaguely imply here.

What point do you think I'm vaguely implying ?

As long as everyone is fully aware that the columns making up each perimeter panel of WTC2 East face were separated by failure of the bolts holding upper and lower panels together, resulting in a staggered break-line across the width of the building. The panel columns themselves sprang back from a state of bowing to their original straight form once the bolts had failed.

 

[BasqueArch]

femr2
Dear me. The assemblies did not buckle. Instead the connection between one assembly and the one above broke along the bolted connections across the width of the building. The columns which make up each of those assemblies did not buckle.

.....
Would you say that "the east wall buckled inward" ?
I wouldn't.

But structural engineers would.
That’s because femr doesn’t know what buckling means. Both the walls and columns buckled.

“There are two main modes of buckling failure that may be experienced by steel members: Overall (or general) buckling and local buckling. Figure 6.1.1 illustrates the difference between the two modes.
General buckling is characterized by a distorted, or buckled, longitudinal axis of the member. In local buckling, the axis of the member is not distorted, but the strength of the cross section is compromised by the buckling of a component of the cross section.”

Figure 6.1.1
General vs. Local Buckling​

http://www.bgstructuralengineering.com/BGSCM/BGSCM006/index.htm

​

The columns buckled (distorted longitudinal axis of the member) whether the buckling was permanent (plastic) or temporary (elastic). The walls buckled (distorted longitudinal axis of the wall) and completely failed at the columns splices (not “seams”) with the splice bolts failing by either punching through or shearing. Columns and walls can buckle without failing to carry the imposed loads. The columns buckled without completely failing, the walls buckled and completely failed at the bolted splices.

NIST :
“With continuously increased bowing and axial loads, the entire width of the east wall buckled inward.
The instability started at the center of the wall and rapidly progressed horizontally toward the sides. As a result of the buckling of the east wall, the east wall significantly unloaded, redistributing its load to the softened core through the hat truss and to the east side of the south and north walls through the spandrels (see Figs. 5–13 and 5–14 and Table 4–38). The section of tower above the buckled wall suddenly moved downward, and the building tilted toward the east (see Fig. 5–15). “


For WTC1,2 the columns buckled, with the walls buckling and completely failing. NIST is right, femr is wrong.

 

[femr2]

For WTC1,2 the columns buckled, with the walls buckling and completely failing. NIST is right, femr is wrong.

LOL. Despite your lengthy (for you) post, with all of it's formatted emphasis, your thinly veiled personal attack is rather moot, as...

I accept the behaviour could be stated as a form of buckling

I accept the behaviour could be stated as a form of buckling.

I have already accepted that fracture of the bolt seams between two separate columns can be termed as a form of buckling, twice.

I note you removed the last quote above from your quote of my text above.

I note also that you have not provided specific answers to...

What governs the direction of the fracture ?

What governs the direction of an elastic buckle ? (around the axis with the lowest moment of inertia I suppose)

Individual direction for each half of the fractured member ? (which in this case could be the bolts...)


The lower columns clearly spring back.


The cracks in the bolts probably formed somewhere between in-plane sliding (placing the propogation direction parallel to the applied shear stress) and out-of-plane sliding (placing the propogation direction normal to the shear stress). More in-plane I suppose ?

Given that the upper column and lower column travel in different directions following failure of the connection between them, what direction would the subsequent buckle be in, and why ?

My original assertions clearly relate to my viewpoint on each column in a panel section being a column in its own right...

I think it's quite natural to describe the columns in each individual perimeter panel assembly as columns in their own right.

Regardless of what other people are saying here, you are correct.

Makes you wonder why NIST would use the term inward bowing, as a separate phase before buckling.

Where does bowing turn into buckling ?

 

[NoahFence]

But structural engineers would.
That’s because femr doesn’t know what buckling means. Both the walls and columns buckled.

Am I missing something in this thread? (Clearly I am....)

I have zero engineering experience beyond my ability to spell the word, and I can see that both the walls and columns buckled as plain as day.

 

[BasqueArch]

LOL. Despite your lengthy (for you) post, with all of it's formatted emphasis, your thinly veiled personal attack is rather moot, as...

A fact is not personal.
Like your it's [sp] is a fact, not personal.

Compare this fact to:

femr2
Please stop this keep-the-faith NISTian-supporter nonsense.

I note you removed the last quote above from your quote of my text above.

I note also that you have not provided specific answers to...

What governs the direction of the fracture ?

What governs the direction of an elastic buckle ? (around the axis with the lowest moment of inertia I suppose)

Individual direction for each half of the fractured member ? (which in this case could be the bolts...)

The lower columns clearly spring back.

The cracks in the bolts probably formed somewhere between in-plane sliding (placing the propogation direction parallel to the applied shear stress) and out-of-plane sliding (placing the propogation direction normal to the shear stress). More in-plane I suppose ?

Given that the upper column and lower column travel in different directions following failure of the connection between them, what direction would the subsequent buckle be in, and why ?

My original assertions clearly relate to my viewpoint on each column in a panel section being a column in its own right...

Makes you wonder why NIST would use the term inward bowing, as a separate phase before buckling.

Where does bowing turn into buckling ?

Ask a structural engineer, seriously. Have you ever discussed your propositions with a SE?

I understand that claiming to still be right while being wrong must be frustrating

 

[femr2]

I understand that claiming to still be right while being wrong must be frustrating

What claim are you referring to ?

I note you have failed to answer my questions...

 

[femr2]

I have zero engineering experience beyond my ability to spell the word, and I can see that both the walls and columns buckled as plain as day.

See here.

Question you could ask yourself are...

Why does NIST use the phrase inward bowing ?

Where does bowing turn into buckling ?

Same thing ? Then why make the distinction ?