Total Building Collapse from a Single Column Failure

[JSanderO]

I assume by "not a single column failure" you mean that the initiator of the collapse was from the failure of A PORTION of the system (IE a group of components failing)? It doesn't really change the final outcome though this and the rest of the post is a hell of a lot more realistic than leaping from "bolts didn't fail" to "fire didn't cause collapse". It may not change the final collapse result, though you could potentially make an argument that if the NIST report detailed the wrong failure point as the collapse initiator there's another concern for life safety that needs to be addressed in newer constructions. I believe with WTC 7 the long-span beams were among chief concerns already and would be a factor in this option as well, but this direction of debate in some cases would shed light on additional code related deficiencies if the ramifications were sufficient.

I'm not sure it affects the conclusions enough but I say this to give you an idea of a legitimate direction of discussion since what you've considered would naturally lead there.

This is the whole discussion about what was the final straw which broke the camel's back... and camel's back here is a huge steel frame. We are told by NIST that the start was on floor 13 col 79... I don't think so. I think that was a consequence of the process begun with the first straw which I suggest was down in the massive load transfer structures.

 

[NoahFence]

Edited by kmortis: 

Removed personal comment

If you really think there is any possibility of a "roll-off" of the girder, I think you should show calculations to back that up. I don't see any possible way for the girder to "roll-off", as the beams and their connections to the girder would have restrained the girder.

Is there a way for us laypeople to get a sort of visual, or maybe just a definition of the term "roll-off"?

 

[LSSBB]

Sure the stiffeners were omitted as a simplifying assumption, and there is a bridge in Brooklyn for sale.

When making simplifying assumptions in structural analyses you aren't supposed to omit anything that would have a significant effect on the results. The stiffeners certainly would have had a very large effect, as in a diametrically opposed conclusion regarding whether or not the girder could have walked off its seat.

It is also customary to sate what you have left out of the model and the stiffeners aren't mentioned anywhere in the report. How come?

Why doesn't Figure 8-21 in NCSTAR 1-9 show them?

You haven't shown that they were significant enough to have been included or even mentioned. You really need to do a full simulation for that, and I'd say laboratory testing also.

 

[NoahFence]

It is clear now that those here who insist that fire caused WTC 7 to collapse have no idea what initiated the collapse with the NIST hypothesis being shown to be impossible, and it is telling that the theories they have proposed to replace the now debunked NIST theory range from barely plausible to ridiculous.

One would think this shows why the NIST should be required to revisit the issue and perform physical testing with sections of floor structure as well as FEA to determine a viable failure mode. Of course, this would also mean they would have to include pertinent structural features they previously omitted, like the girder stiffeners.

It HAD to be fire. There is no other physically possible mechanism that could have taken down that building, at that time, under those circumstances.

When implying controlled demolition, why are you not taking into account the very real implications of that? Human intervention, timing, scope*, raging fire - all these need to be overcome in order for controlled demolition to work.

The other option, fire, will also do the job if you just sit there and stare at the building long enough, without so much as a cup of coffee being spilled on the fire to fight it.

One of the two above paragraphs actually happened. Which one?

Reminder - if you chose the first one, you need to explain the issues I've presented.


* Scope -
This would be almost double the previous world record demolition. Think about that.

 

[tsig]

Oh, but the calculations do indeed show those stiffeners would have prevented the girder flange from failing when the web was past the seat, and by a large margin at elevated temperature also. There is a reason they were omitted and not mentioned in the report and it wasn't to simplify the analysis.

Can I see your calculations which show otherwise? Or even a little logic if you can't do the calculations?

Here is some logic.

1) the building had an unfought fire for seven hours

2) fires are hot

3) heat weakens steel

Conclusion: the heat from the fires weakened the steel below the load it had to support.


Truther logic.

1) Conclusion: it was a CD. The US govmt* is evil.

2) search for anything to try and prove 1



*insert NWO, Jews ect.

 

[Animal]

There's this post, which you handwaved away with the aforementioned "blah blah blah" that I've come to expect from you. If you gave an actual reasoned response to it instead of just treating it like technobabble you are unable to understand, that would be a great start.

When TS gets boxed in...he returns to linear thinking and the "pristine building" fantasy

 

[Grizzly Bear]

This is the whole discussion about what was the final straw which broke the camel's back... and camel's back here is a huge steel frame. We are told by NIST that the start was on floor 13 col 79... I don't think so. I think that was a consequence of the process begun with the first straw which I suggest was down in the massive load transfer structures.

Honestly im not as much into that level of detail as you may be so I cant say if I disagree on this or not in particular. However, as I said, carried out the right way, you have room to argue it if no where else on the grounds of building code changes that would probably be affected. In other words the level of detail dictates CAN have relevance. It just doesnt work out in tonys case because whereas you can make a justification on relevant code changes to critique the nists conclusiins, tony is trying to link these claimed errors to a conclusion that has absolutely no basis in reality.

That demonstrates exactly how important starting context and positions are in the context of this discussions and makes the difference between what argunents hold water versus those that are distractions

 

[BasqueArch]

According to your calculations, the girder buckled, not deflected 2-3" like you claim. It's all over then.
Also the beams would have buckled when hot and having lost their stiffening shear bolts. Girder expanded, jammed at colums, resisted beam expansion, added axial resistance, more buckling, loss of load carrying geometry. Horizontal catenary forces on girder. Also over.
Girder or beams failed first.
Buiding collapsed due to the fires not controlled demolition.

The five beams won't let the girder rotate, and it would need to rotate to lose load carrying capacity. So this failure mode is unlikely.

What is interesting is that you at least acknowledge that the NIST hypothesis is impossible.

What is interesting is that you finally acknowledge that the 44-79 girder buckled.

And no, the NIST theory that fires caused the collapse is the only possible one. CD and Oystein's Midgets with Saws are possible, but since there is zero evidence for either, epistemology requires that fire is the only possible explanation.

NIST analyzed various fatal beam-girder failure modes including the push off. The girder stiffener at floor 13 is not proof that the web once past the seat would prevent the girder from rotating off (NB). Some of the other girders on other floors at column 79 did not have stiffeners. Also the coefficient of expansion increases as the temperature increases. The 16 floor model used the Fire Dynamics Simulator and produced girder temperatures of up to 1000C. Also the small heated seat, with the sagged girder's weight on one corner of it would have bent or sheared it; however NIST chose not to analyze this.

The CD hypothesis is impossible.

 

[Newtons Bit]

Addressing a claim

The 53 foot long beams to the east of girder A2001 would buckle at approximately 8,000 lbs of axial force when they were at 600 degrees C. The six 7/8" diameter ASTM A325 bolts in their connection to the girder would require about 18,000 lbs. each to shear at 600 degrees C. That is 108,000 lbs. to break the bolts at 600 degrees C and 8,000 lbs. to buckle the beam.

The above is one of the argument Tony Szamboti is making for a "pristine" building. The beam in question is a 52ft long W24x55. His claim is not true, and I'll explain why.

Background
Axial buckling occurs when the axial compressive stress in a member exceeds the critical buckling stress (F_cr). F_cr is a function of the unbraced length in either the major axis, or the minor axis of the member. For a typical floor beam, the vertical direction is the major axis (aka the strong axis) while the horizontal direction is the minor axis (aka the weak axis). A typical floor beam typically has no bracing in the major axis (except at the ends) and it is braced in the minor axis by beams framing in to it. For example, the girder A2001 that spans between column 79 and column 44 is braced by the multiple beams that frame into it. The weak axis is also considered continually braced by headed anchor studs that are part of a composite floor deck (see AISC 360-10 Commentary I7 (p. 16.1-375)).

As the beam begins to heat, it expands. The shear studs restrain this expansion. At about 100C, the first shear studs begin to fail. By 300C, all but 3 shear studs have failed (see NCSTAR 1-9 p.353). In between these two events, the beam is expanding in a partially restrained condition and loading the bolts at the end. Many of the shear studs are still intact, these shear studs brace the beam against minor axis buckling.

Per NIST, the nominal capacity of each headed anchor stud is about 20kip and there are 28 of them per beam (roughly 2ft on center).

Thermal Strain
At 100C, the beam has a thermal strain (d_T) of 0.00098 (see NCSTAR 1-9 p.344) and at 300C, the beam has thermal strain of 0.00378. From simple engineering mechanics, strain is:

d= P/AE

or

P = d_T * A * E

Where A is the cross-sectional area of the member, E is the Modulus of Elasticity and P is the axial force required to restrain the above strain.

For the W24x55 beam:

P(dT = 100C) = 0.00098 * 16.2in² * 29000ksi = 460kip
P(dt = 300C) = 0.00378 * 16.2in² * 29000ksi = 1776kip

Compressive Strength
F_cr can be directly computed (or looked up in a table as I'll do) but the member's slenderness ratio:
K*L / r

where K = 1.0 for our case, L is the unbraced length of the member in the major or minor axis and r is the radius of gyration (a derived unit) in the respective axis.

Let's assume that many, but not all, of the headed anchor studs have failed at some temperature between 100C and 300C. The unbraced length of the major axis is still the entire length of the beam. The unbraced length in the minor axis is the distance between these studs. If we were to assume that there were 7 shear studs left, the unbraced length for minor axis buckling would be somewhere around 8ft.

KL/r (major) = 1.0 * 624in / 9.11in = 68.4
KL/r (minor) = 1.0 * 96 / 1.34in = 71.6

The larger slenderness ratio, 71.6, controls. AISC 360-05 table 4-22 (aside: this table was removed in AISC 360-10!) shows the F_cr of a Fy = 50ksi w/ a slenderness ratio of 72 as 30.8ksi. This includes a phi factor of 0.9, removing that gives an F_cr of 34.2ksi. The axial capacity of the W24x55 beam is:

P_n = A*F_cr

P_n = 16.2in²*34.2ksi = 554kips.

Conclusion
A few drinks later and its time to dot our t's and cross our i's. We need to acknowledge that we don't know what temperature the shear studs are in the assumed condition. But we do know that is at least 100C. At that temperature, the beam could impart up to 460kip on the beam-girder connection. We also know that the 7 remaining shear studs could reduce that axial force of 460kip by 20kip each. There's also a ~20kip vertical shear reaction from gravity loads at the connection. In total, that means there's about 340kip available to damage the beam-girder connection, a fair bit higher than the 8kip Tony Szamboti has claimed.

tl;dr: Tony Szamboti is wrong. This is really complicated and time dependent, use a 3D model for best results.

 

[Grizzly Bear]

Just out of curiosity, and heavily dependent upon the amount of info available about the building but would there be any chance that some limited analysis could be done with a revit model? This wouldn't be for any kind of fire analysis, but to study the design parameters better is why I ask... I've considered modeling the structure someday when I'm dead beat bored out of my wits to actually look at some of this for my own purposes.

 

[Tony Szamboti]

The above is one of the argument Tony Szamboti is making for a "pristine" building. The beam in question is a 52ft long W24x55. His claim is not true, and I'll explain why.

Background
Axial buckling occurs when the axial compressive stress in a member exceeds the critical buckling stress (F_cr). F_cr is a function of the unbraced length in either the major axis, or the minor axis of the member. For a typical floor beam, the vertical direction is the major axis (aka the strong axis) while the horizontal direction is the minor axis (aka the weak axis). A typical floor beam typically has no bracing in the major axis (except at the ends) and it is braced in the minor axis by beams framing in to it. For example, the girder A2001 that spans between column 79 and column 44 is braced by the multiple beams that frame into it. The weak axis is also considered continually braced by headed anchor studs that are part of a composite floor deck (see AISC 360-10 Commentary I7 (p. 16.1-375)).

As the beam begins to heat, it expands. The shear studs restrain this expansion. At about 100C, the first shear studs begin to fail. By 300C, all but 3 shear studs have failed (see NCSTAR 1-9 p.353). In between these two events, the beam is expanding in a partially restrained condition and loading the bolts at the end. Many of the shear studs are still intact, these shear studs brace the beam against minor axis buckling.

Per NIST, the nominal capacity of each headed anchor stud is about 20kip and there are 28 of them per beam (roughly 2ft on center).

Thermal Strain
At 100C, the beam has a thermal strain (d_T) of 0.00098 (see NCSTAR 1-9 p.344) and at 300C, the beam has thermal strain of 0.00378. From simple engineering mechanics, strain is:

d= P/AE

or

P = d_T * A * E

Where A is the cross-sectional area of the member, E is the Modulus of Elasticity and P is the axial force required to restrain the above strain.

For the W24x55 beam:

P(dT = 100C) = 0.00098 * 16.2in² * 29000ksi = 460kip
P(dt = 300C) = 0.00378 * 16.2in² * 29000ksi = 1776kip

Compressive Strength
F_cr can be directly computed (or looked up in a table as I'll do) but the member's slenderness ratio:
K*L / r

where K = 1.0 for our case, L is the unbraced length of the member in the major or minor axis and r is the radius of gyration (a derived unit) in the respective axis.

Let's assume that many, but not all, of the headed anchor studs have failed at some temperature between 100C and 300C. The unbraced length of the major axis is still the entire length of the beam. The unbraced length in the minor axis is the distance between these studs. If we were to assume that there were 7 shear studs left, the unbraced length for minor axis buckling would be somewhere around 8ft.

KL/r (major) = 1.0 * 624in / 9.11in = 68.4
KL/r (minor) = 1.0 * 96 / 1.34in = 71.6

The larger slenderness ratio, 71.6, controls. AISC 360-05 table 4-22 (aside: this table was removed in AISC 360-10!) shows the F_cr of a Fy = 50ksi w/ a slenderness ratio of 72 as 30.8ksi. This includes a phi factor of 0.9, removing that gives an F_cr of 34.2ksi. The axial capacity of the W24x55 beam is:

P_n = A*F_cr

P_n = 16.2in²*34.2ksi = 554kips.

Conclusion
A few drinks later and its time to dot our t's and cross our i's. We need to acknowledge that we don't know what temperature the shear studs are in the assumed condition. But we do know that is at least 100C. At that temperature, the beam could impart up to 460kip on the beam-girder connection. We also know that the 7 remaining shear studs could reduce that axial force of 460kip by 20kip each. There's also a ~20kip vertical shear reaction from gravity loads at the connection. In total, that means there's about 340kip available to damage the beam-girder connection, a fair bit higher than the 8kip Tony Szamboti has claimed.

tl;dr: Tony Szamboti is wrong. This is really complicated and time dependent, use a 3D model for best results.

For stress to occur in the bolted connections between the beams and the girder, as you want to claim, there not only needs to be force from the beams (which there certainly is) but there needs to be a reaction from the girder. The girder had no shear studs, so its only reaction was due to friction and that has been shown to be negligible relative to the strength of the connections.

Everything you said here is moot due to a lack of sufficient reaction from the girder.

 

[Newtons Bit]

Just out of curiosity, and heavily dependent upon the amount of info available about the building but would there be any chance that some limited analysis could be done with a revit model? This wouldn't be for any kind of fire analysis, but to study the design parameters better is why I ask... I've considered modeling the structure someday when I'm dead beat bored out of my wits to actually look at some of this for my own purposes.

Revit itself cannot do any structural analysis. It can plug directly into Robot (or so Autodesk claims), but I've never had any luck with that product. You'd need to use a product like RISA, STAAD or something similar.

ETA: Maybe I misunderstood you. Are you looking to create a non-analytical model for visual purposes?

 

[MileHighMadness]

Gee...if Tony didn't have this forum, who would he talk to? ��

 

[Newtons Bit]

Gee...if Tony didn't have this forum, who would he talk to? ��

A fair point.

 

[ozeco41]

Gee...if Tony didn't have this forum, who would he talk to? ��

In recent months he has been visiting at least one other forum which lacks the level of engineering understanding the members here possess - "jointly and severally".

Between us we have a wide ranging coverage of the topics.

 

[Grizzly Bear]

ETA: Maybe I misunderstood you. Are you looking to create a non-analytical model for visual purposes?

I was hoping revit would have the capacity to run very basic load analysis based on BIM. I would have used the information to see how loads were distributed in the as-built building. Nothing on the level of what NIST does of course.

 

[funk de fino]

Has TS found that video that he claimed to see yet?

 

[Axxman300]

The other factor Truthers ignore is that WTC7's fires began after parts of WTC1 crashed into it. The initial damage has never been completely revealed, most due to the lack of objective thinking after the second collapse as FDNY, and everyone else launched into full rescue mode.

So we don't know what kind of wreckage struck WTC7, nor the penetration of this wreckage. However, it is a safe assumption that heavy steel beams ended up inside. Even one changes the structural load on the floor it landed on, and if this beam were hot (and started the fire) it would be a significant factor in the collapse.

The key is that WTC7 went down as a result of damage from WTC1, and that's where this discussion starts. Truthers ignore this, and none of their homework addresses the initial damage.

 

[Newtons Bit]

I was hoping revit would have the capacity to run very basic load analysis based on BIM. I would have used the information to see how loads were distributed in the as-built building. Nothing on the level of what NIST does of course.

Nope. Revit has no built-in structural analysis tools. You may have been mislead by all the analytical parameters for structural families. Those are there to tell export/import tools how you want certain elements treated. You'll need something like Robot or RISA Floor (both of which tie into Revit, the latter with supposedly great ease).

 

[Tony Szamboti]

The other factor Truthers ignore is that WTC7's fires began after parts of WTC1 crashed into it. The initial damage has never been completely revealed, most due to the lack of objective thinking after the second collapse as FDNY, and everyone else launched into full rescue mode.

So we don't know what kind of wreckage struck WTC7, nor the penetration of this wreckage. However, it is a safe assumption that heavy steel beams ended up inside. Even one changes the structural load on the floor it landed on, and if this beam were hot (and started the fire) it would be a significant factor in the collapse.

The key is that WTC7 went down as a result of damage from WTC1, and that's where this discussion starts. Truthers ignore this, and none of their homework addresses the initial damage.

Anyone should be able to realize that at least half of WTC 1 fell on WTC 7. WTC 1 wreckage had long legs (as WTC 7 was 350 feet away) and it was all hot and capable of starting fires. It is a wonder only ten floors in WTC 7 had fires started in them.