Moderated Continuation - Why a one-way Crush down is not possible

[R.Mackey]

So if I'm reading this correctly, I don't think they're actually commenting that the majority of connections were moment connections. I think they're just discussing what elements they included in a modeling of a specific element. Ryan, Newton's, have I got that right? Feel free to correct me if I'm wrong.

That's correct. They note there were moment connections -- as mentioned previously, on the beam-framed floors, and also where there were gaps in the core flooring as a way to replace lost diaphragm stiffness. But such moment connections are far and away the exception.

In the ANSYS modeling, they didn't include non-moment connected beams at all, instead replacing with an idealized diaphragm. Only the special, rare beams were included as beams in that model.

If NIST actually said what Tony claimed they said, i.e that most of the beams were welded to take moment, it would have been in NCSTAR1-1, not 1-6D. That's your first clue.

 

[Tony Szamboti]

Um, dude...
Strain is the INDEPENDENT AXIS in those graphs.
Stress is a function of strain, not the other way around...
And if you'll note, the curve is really, really flat out there in the plastic region.
Which means that Young's Modulus (Slope of the curve) out there is effectively zero, so that if you consider a beam in bending,
Y=F*L^3)/(48*E*I)
where Y is deflection
F=applied force
L=length between supports
E=Young's Modulus
I= Moment of Inertia--see Newtons Bit' post

What happens when E goes to zero, there, dude?

We are talking about axial strain due to compression here not bending.

Anybody who understands strain hardening knows that to get additional strain in the plastic region requires a higher stress which would require an additional load and since

axial deflection = PL/AE

if the load (P) could not increase, and increase the stress, then the modulus (E) had to decrease, which could only happen with an increase in temperature.

I understand how to calculate the moment of inertia of a beam or column.

 

[Tony Szamboti]

Tony,

This conversation is starting to get surreal...

You published a paper with Mr. Legge. I just became aware of your interpretation of a particular graph.

I asked if you stood behind that paper. You replied:



I was frankly puzzled, because there is absolutely nothing in that paper that makes this claim. I answered you directly, [ http://www.internationalskeptics.com/forums/showthread.php?postid=4983748#post4983748 ] that "No, that not what I am saying & that's not what you said."

Then I addressed - directly - several of the comments that you made that I believe (OK, you got me... "that I know") to be incorrect. In order to be precise with your statements, I addressed them by cutting & pasting your words verbatim. And then asserted that several of your comments were wrong, and explain exactly why they are (IMO) wrong.

My comments were:

1. Why did you use curves for 43A instead of A36 steel? (Since the towers were made with A36.)

2. The correct interpretation of those curves shows the decrease in yield strength, ultimate strength & elastic modulus with rising temp.

3. Bulk modulus cross sectional expansion is irrelevant once the columns go into bending. (Others have shown it to be inconsequential, even if the column stresses stay purely compressive.)

4. Your comment that "As the steel sags ... the inherent strength of the steel will increase" is totally, utterly wrong.

5. Your comment that "the yield strength of steel increases as the degree of distortion increases" is totally wrong.

6. Since the effect that you claim in 5. above is false, it cannot, as you further claim, become more pronounced at elevated temperatures.

7. Your claim that "initial sag in steel cannot be catastrophic..." is wrong.

8. Your claim that "A rising temperature will be needed to offset both the significant increase in yield strength and the slight increase in cross-section area, if collapse is to progress" is completely wrong.

9. You comment that "the upper section should only have moved down slowly and only continued to do so if additional heat was supplied" is both unproven by anything that this paper offers and completely wrong to boot.

___

And now, you AGAIN come back with this reply...



Tony, I don't get it. You're clearly not dumb. You don't strike me as being intentionally, insultingly rude like Heiwa.

Why do you keep suggesting that I've said ANYTHING about energy added after buckling??

Tom

PS. BTW, just to be complete with you, I do believe that you have to add mechanical energy to continue to collapse a beam after it buckles. A small, trivial amount of energy compared to the amount to bring the beam up to the point of buckling.

As described by the correct graph (on the left below).

[qimg]http://www.internationalskeptics.com/forums/picture.php?albumid=176&pictureid=1437[/qimg]

The energy required is, as indicated on this graph, the area under the Force vs. deflection curve.

And, as shown, FAR from getting stronger, the beam gets much weaker after buckling.

Tom

The paper discusses failure due to compressive rupture, not buckling. Why would you be discussing buckling here? It sounds like you are intentionally distorting what was said in the paper and I really don't want to play your game.

Steel does strain harden once it is yielded and in the plastic region. Additional energy is required to cause additional strain.

 

[ElMondoHummus]

That's correct. They note there were moment connections -- as mentioned previously, on the beam-framed floors, and also where there were gaps in the core flooring as a way to replace lost diaphragm stiffness. But such moment connections are far and away the exception.

In the ANSYS modeling, they didn't include non-moment connected beams at all, instead replacing with an idealized diaphragm. Only the special, rare beams were included as beams in that model.

If NIST actually said what Tony claimed they said, i.e that most of the beams were welded to take moment, it would have been in NCSTAR1-1, not 1-6D. That's your first clue.

Yes, exactly. 1-1 actually discussed the design and construction. 1-6D was about the analysis, which of course was an extended discussion of the computer modeling.

 

[Newtons Bit]

NIST NCSTAR 1-6D pages 168 and 169

There were a grand total of two beams with moment connections. This makes the two individual frames* stronger in one direction, but just as weak in the other. It doesn't change anything. There still wasn't lateral stability in the core without the diaphragm connection to the perimeter moment frames.

*- The frames are just two columns.

 

[Grizzly Bear]

The paper discusses failure due to compressive rupture, not buckling. Why would you be discussing buckling here? It sounds like you are intentionally distorting what was said in the paper and I really don't want to play your game.

Steel does strain harden once it is yielded and in the plastic region. Additional energy is required to cause additional strain.

I would probably be discussing buckling failure modes because that was how the columns in the impact regions failed. I'm still confused why compressive rupture required the degree of attention dedicated to it in that paper if it was not the primary culprit of the collapse initiation. I admit I was a bit confused about what failure modes you and Legge were talking about in that excerpt ("sagging" wasn't a particularly clear terminolgy), but it kind of concerns me even more that it took on this tangent.

 

[Gravy]

Yes, exactly. 1-1 actually discussed the design and construction. 1-6D was about the analysis, which of course was an extended discussion of the computer modeling.

Typically, Szamboti ignores the obvious and convinces himself that his reality applies to the world outside his mind. He somehow forgets that we know how the towers were constructed. I cannot fathom what processes go on in the labyrinths of his mind that make him think he can convince us that the towers were constructed in some other way, but there you have it.

 

[R.Mackey]

What I can't fathom is why he thinks we'd be so easily fooled. Anyone can read the reference. Its meaning is quite clear, and it certainly says nothing even resembling "most of the central core beam to column connections were moment connections. "

As I have before, I strongly advise him to do better when he debates me on Hardfire. If he's going to make mistakes like this, it's going to be over quickly.

 

[Tony Szamboti]

I would probably be discussing buckling failure modes because that was how the columns in the impact regions failed. I'm still confused why compressive rupture required the degree of attention dedicated to it in that paper if it was not the primary culprit of the collapse initiation. I admit I was a bit confused about what failure modes you and Legge were talking about in that excerpt ("sagging" wasn't a particularly clear terminolgy), but it kind of concerns me even more that it took on this tangent.

The columns had low slenderness ratios so buckling as the initiating mechanism was unlikely.

 

[Tony Szamboti]

Typically, Szamboti ignores the obvious and convinces himself that his reality applies to the world outside his mind. He somehow forgets that we know how the towers were constructed. I cannot fathom what processes go on in the labyrinths of his mind that make him think he can convince us that the towers were constructed in some other way, but there you have it.

It is starting to look like you CAN'T address the topic of this thread as requested by Heiwa, the thread originator. If you can't then what are you doing here?

 

[R.Mackey]

... mistakes like that, for instance ...

 

[Grizzly Bear]

The columns had low slenderness ratios so buckling as the initiating mechanism was unlikely.

It was certainly observed in the perimeter columns both before & during the initiation of the south tower's collapse (I added some notation to these):

Same for the north tower

It is starting to look like you CAN'T address the topic of this thread as requested by Heiwa, the thread originator. If you can't then what are you doing here?

Maybe it hasn't occurred to you that everything you write has consequences to your credibility:

Like the things you've said about the debris plumes during collapse.

Try to explain their upward movement with a gravity only collapse. They are narrow and appear to emanate from point like sources not a pressure wave. When standard controlled demolitions are done and the building falls to the ground the cloud is wide and diffuse not narrow..

Epic mistakes... do a number.... on credibility... I've also been a little curious if you still stood by this as well.

 

[Tony Szamboti]

There were a grand total of two beams with moment connections. This makes the two individual frames* stronger in one direction, but just as weak in the other. It doesn't change anything. There still wasn't lateral stability in the core without the diaphragm connection to the perimeter moment frames.

*- The frames are just two columns.

Where do you get that just two beams had moment connections?

There is a plan view figure on page 169 that shows which connections were included, as they were moment connections, and there are a lot more than two. Judging from that figure I would say 70% were moment connections.

I have said enough here and I am being tempted to respond to Ryan Mackey and I said I would not do that until after the debate. So it is Ciao for now.

 

[rwguinn]

We are talking about axial strain due to compression here not bending.

Anybody who understands strain hardening knows that to get additional strain in the plastic region requires a higher stress which would require an additional load and since

axial deflection = PL/AE

if the load (P) could not increase, and increase the stress, then the modulus (E) had to decrease, which could only happen with an increase in temperature.

I understand how to calculate the moment of inertia of a beam or column.

You obviously have no clue what "E" is, however...

 

[R.Mackey]

Tony, the plan view shows there were two. They're the ones indicated by the arrows.

You read the figure wrong.

This is not a worthy topic for debate, so go ahead and reply.

 

[tsig]

<munching more popcorn, this is sooooooo much better than G.I. Joe <shudder>. Munch munch munch</munching>

Heiwa claiming to be an engineer is like someone claiming to be a race driver but running into the wall every time he starts the engine.

 

[Tony Szamboti]

You obviously have no clue what "E" is, however...

E = stress/strain

 

[ElMondoHummus]

Tony, you're also missing the point that was made above, that NCSTAR 1-6D isn't a report on the actual construction of the towers. It's a discussion on the modeling, which includes the choices and tradeoffs they made in creating the computer model. They stated earlier that they only included the moment connections in the modeling and did something else to represent the other connections.

 

[twinstead]

It is starting to look like you CAN'T address the topic of this thread as requested by Heiwa, the thread originator. If you can't then what are you doing here?

The only one who can't address the topic of the thread is Heiwa, the thread originator. That's the whole issue.

 

[Tony Szamboti]

Tony, the plan view shows there were two. They're the ones indicated by the arrows.

You read the figure wrong.

This is not a worthy topic for debate, so go ahead and reply.

I interpreted the figure as those that were shown being moment connections as there are several left out of the figure.

If the arrows pointing to the two beams are intended to show that they are the only moment connections then I stand corrected. However, I would add that the discussion of which were and weren't moment connections could be clearer and I don't know why it isn't and why the NIST does not fully describe the connections of all of the beams. I think you have to admit that they don't do that.