How Buckling Led to "Free Fall" acceleration for part of WTC7's Collapse.

[Jeffrey Orling]

The primary mechanism of the initiation stage of the Twin Towers collapses was a sequenced cascading failure of columns losing their capacity to carry vertical loads. That process does NOT require: "...that the collapse would require almost all the trusses to behave like that in the same time frame..." NOR does it require: "...that the fires were throughout the footprint over entire floor area.." So your conclusion is moot (irrelevant) and not supported.

OK... My rejection of NIST's sagging truss pull in causing buckling of the exterior of the twin towers includes:
they provided evidence of this at only one location on the east side of the south facade on 1WTC. Cameras were trained on the towers continuously after the plane strikes. No evidence of large portions of the facade being pulled in or "buckling" in either tower

If the trusses were to pull in the facade as a result of their being heated and sagging... one would think that there would be excessive heating over all or most of the length of the truss for the entire truss to sag. If this is true it would mean that the entire outside for core floors were engulfed in flames. There is no photographic evidence of this.

The pull in would be 1/2 as much on the short span trusses (not that it matters)

It makes sense that the fuel from the plane would not be distributed equally throughout the foot print

Considering 2WTC almost all fuel would be in the SW and little no not on the NW side of the tower. In 1WTC it appears that the fuel would mostly be in the core area where is could enter and fall down elevator shafts. NIST's pull in tower 1 was the furthest from the plane strike. How did that work?

Econ41 correctly explains how a progression of column failures likely occurred in the core region and load redistribution could drive overall capacity below service loads leading the the tops dropping and tilting and the top blocks sheared floors, crushed floors and became the mass which in a runaway fashion rapidly destroyed the floors sequentially from the plane strike zone down the the ground.

For some reason NIST ignored the most logical hypothesis for the tops dropping... lost of column capacity in the core.... and choose an illogical one of the perimeter buckling.

 

[econ41]

OK... My rejection of NIST's sagging truss pull in causing buckling of the exterior of the twin towers includes:
they provided evidence of this at only one location on the east side of the south facade on 1WTC. << Correct Cameras were trained on the towers continuously after the plane strikes. No evidence of large portions of the facade being pulled in or "buckling" in either tower << Correct

The mechanism was a "sequenced cascading failure of columns losing their capacity to carry vertical loads. That process does NOT require: "...that the collapse would require almost all the trusses to behave like that in the same time frame..." NOR does it require: "...that the fires were throughout the footprint over entire floor area.." So your conclusion is moot (irrelevant) and not supported.

If the trusses were to pull in the facade as a result of their being heated and sagging... one would think that there would be excessive heating over all or most of the length of the truss for the entire truss to sag.

No "one would think" that if one was thinking clearly. A cascade failure is sequential. It starts from one trigger point. It does NOT require every column to be heated, buckled OR inwards bowed.

If this is true it would mean that the entire outside for core floors were engulfed in flames.

It isn't true no matter how many times you repeat it.

There is no photographic evidence of this.

Of course. It didn't happen.

Econ41 correctly explains how a progression of column failures likely occurred << Correct so far in the core region << Wrong .......

I have NEVER limited the process of load redistribution to core only. There is no magic way for core only redistribution to occur. It MUST involve both core and perimeter.

For some reason NIST ignored the most logical hypothesis for the tops dropping... lost of column capacity in the core....

NIST was under no obligation to comply with your preferred explanation. It is not the most logical hypothesis. It may be one plausible hypothesis. But it still cannot go against the 3 realities that (1) load redistribution occurred, (2) Load redistribution does NOT exclude either perimeter or core columns from the process AND (3) load redistribution by its very nature starts from a trigger point..

and choose an illogical one of the perimeter buckling.

Well NIST got it right.

 

[Jeffrey Orling]

The mechanism was a "sequenced cascading failure of columns losing their capacity to carry vertical loads. That process does NOT require: "...that the collapse would require almost all the trusses to behave like that in the same time frame..." NOR does it require: "...that the fires were throughout the footprint over entire floor area.." So your conclusion is moot (irrelevant) and not supported.

No "one would think" that if one was thinking clearly. A cascade failure is sequential. It starts from one trigger point. It does NOT require every column to be heated, buckled OR inwards bowed.

It isn't true no matter how many times you repeat it.

Of course. It didn't happen.

I have NEVER limited the process of load redistribution to core only. There is no magic way for core only redistribution to occur. It MUST involve both core and perimeter.

NIST was under no obligation to comply with your preferred explanation. It is not the most logical hypothesis. It may be one plausible hypothesis. But it still cannot go against the 3 realities that (1) load redistribution occurred, (2) Load redistribution does NOT exclude either perimeter or core columns from the process AND (3) load redistribution by its very nature starts from a trigger point..

Well NIST got it right.

I am not going to debate, nor be lectured to.
I express what appears sensible to me based on what I observed... especially regard the buckling of the perimeter.
I saw evidence of that in one location BEFORE the downward movement of the tops.
I see no reason for the perimeter columns to buckle as a result of being overloaded
I see no evidence of floor wide fires which could cause an all around perimeter pull in.
I assume perimeter buckling could result from loss of lateral bracing from truss failure

I can't see inside the core.... no one could.

The core driven explanation is ah hypothesis which assumes:
destruction of some of the core perimeter columns by the plane impact
extensive fires centered in the plane's path through the towers.... Planes broke apart and parts and fuels with momentum continued into the building
Liquid fuel would disburse throughout the floor as spreads... and spread the fires
Lateral steel in the core was likely weakened and lengthened from the heat
Expanded lateral steel in the core would fail to brace the core columns
Expanded lateral steel in the core might force column ends to mis align leading to buckling
Columns lost strength from heat
Columns lost strength when the bracing was non performing or destroyed.
Core loads from failed columns were redistributed by intact lateral steel above the strike zone and the intact hat truss
The hat truss was largely supported by the core
The hat truss acted like a rigid end plate on a hollow tube
The hat truss may have distributed perimeter core column loads to the perimeter columns
23 Interior core columns supported significantly lower loads than 24 core perimeter columns
Core side of the floor slabs were supported on the belt girder which was cantilevered on beam stubs from the 24 perimeter columns of the core.

NIST got some things correct and ignored many other things. Not quite GIGO but it looks more like an conclusion in search of an explanation.

 

[Jeffrey Orling]

My sense is that for the perimeter columns to buckle such that there would be inadequate support for the entire top block a large number of the panels would have to fail.
Remember also the panels were staggered and were 3 stories in height and the floor connections to the columns was not at the column ends but at spandrel height/locations at every other column.

Also to pull in the facade panels would mean crushing the slabs which were like rigid plates holding stiffening the exterior tube. That seems a stretch.
Any buckling of the facade likely was from load redistribution and buckling.
But what we saw was top block lateral movement.

 

[Jeffrey Orling]

@econ41 I am perfectly aware that there can be a progression of column failures in stone cold steel frames. Once a single column fails for whatever reason... the load it had carried must find new load paths and those new paths are to adjacent intact columns. This may overload those columns and the process can repeat and go runaway... taking out one column after another.

In the case of the twin towers, the high loaded and very strong 24 perimeter core columns were among the first "failed" columns caused by the plane impacts. We are told that there were core columns which were not severed by dented or bent and made less stable and more prone to buckling. And we know that the interior core columns were not as highly loaded as the perimeter... and likely had less capacity to sustain distributed loads.

It seems plausible that a progression of column failures could progress through the core which not only defined the elevator shafts but supported the hat truss which play a key role in the towers' rigidity.

 

[econ41]

I am not going to debate,

Then please stop posting Gish Gallops and then ignoring my reasoned and correct responses.

IF you don't want to debate then stop posting false claims which have been rebutted many times and which simply beg to be rebutted again.

Please stop derailing the thread.

 

[econ41]

@econ41 I am perfectly aware that there can be a progression of column failures in stone cold steel frames. Once a single column fails for whatever reason... the load it had carried must find new load paths and those new paths are to adjacent intact columns. This may overload those columns and the process can repeat and go runaway... taking out one column after another.

That is the process we are discussing and you have just made two lengthy posts denying.

It seems plausible that a progression of column failures could progress through the core which.....

Including all the core and perimeter columns as I have explained many times.

 

[Jeffrey Orling]

I have never denied that when a column fails, its loads are taken up by remaining columns usually the closest ones.

The facade had an effective "structure" for load distribution being panels of 3 columns and 3 spandrels as a composite structure. It behaved structurally like a membrane.

 

[econ41]

Let me take the next step of productive discussion with @Thomas B (and @Henkka if he is interested. @Thomas B asked:

What's the next step?

And I said

1) A bit of revision - wait for my next post.

These explanatory diagrams arose in the setting of discussions about 10 or 12 years ago. When The common false mantras (premises) of the truth movement were becoming embedded. e.g. "Free Fall", "fall in own footprint" and "symmetrical" all assumed to somehow "prove" CD. They don't but still serve as the foundations of mistaken belief for gullible followers.

At a slightly higher level of physics was the belief that if a proportion of columns was removed from a column array the load those columns previously carried would transfer uniformly to other columns. If 25% of columns were removed the result would be an additional 33.3% load uniformly applied to the remaining columns.

One version of the fallacy would claim "With only 15% of columns damaged that leaves 85% - more than enough to carry the load since there is an "x" times factor of safety.

Nothing is further from the truth. Where the load goes in such a situation depends very much on the column layout and the locations from which columns were removed.

But truthers were relying on the argument and, sadly, many debunkers did not recognise the error, Including engineering qualified debunkers who should have known better.

So I decided to do a bit of explaining. And the first graphic I produced was a deliberately over-simplified resemblance to a WTC Twin Tower. Three rows of columns and we cut one row. Here it is with the original loads:

So IF we follow thew truther belief "loads redistribute equally" The R side 100 load would go 50 each to the other two columns. (OR 33 - 66 if we follow a slightly different interpretation of the false understanding.)

What really happens (I'm assuming a rigid Top Block) is this:

So, with that example - a doubling of load on one column NOT the lesser uniform proportion.

So that much will be obvious to you @Thomas B and our other two currently active members whu are seeking explanations.

The "Seven Columns" model was the upper limit of complexity when trying to explain to persons who fell for the "Uniform redistribution" falsehood. It is the starting point for the level of complexity we need to understand as we move forward.

Here again is the"7 Columns" model - cut columns version:

Assuming all columns initially had equal loads of (say) 100 THEN cutting "A" and "B" would release 200 for re-distribution AND the false "uniform redistribution" model would see an additional load of 40 applied to Columns C, D, E, F and G. Well within the typical 3 times factor of safety.

What really happens?

The downwards force of 100 at each of A and B will attempt to tilt the top beam by pivoting over "C". It gets a bit complicated to derive actual values for several reasons. However, with C acting as a fulcrum the left end of Top Beam will tend to move down resulting in lifting off D through G And the load on C will increase by more than A plus B because of the leverage of the A>C and B>C moment arms.

That is a simplified overview - the reality is a bit ( ) more complex. BUT it should be easy to see that Column C could easily fail in the real explanation whilst it would barely be overloaded in the Truther advocated "uniform redistribution" false scenario.

So that was a somewhat lengthy explanation of the history of the misunderstanding. And an outline of the type of mechanism which can trigger the "first column to fail" which is necessary to start a cascading failure. Which is the type of failure that would have led to the FFA of the WTC7 perimeter shell facade. We are on topic - explaining the necessary foundations of physics needed to answer the OP Topic.

Always providing there was no CD which could result in simultaneous failure. Proof of CD is a separate topic.

 

[econ41]

I have never denied that when a column fails, its loads are taken up by remaining columns usually the closest ones.

Please @Jeffrey - if you want to disagree with anything I claim then disagree with what I claim...Not your own strawman which I did not claim.

 

[Jeffrey Orling]

sure... how do you see failures progressing in the twin tower panels which had 3 columns and 3 spandrels?

 

[econ41]

sure... how do you see failures progressing in the twin tower panels which had 3 columns and 3 spandrels?

Remember that "Twin Towers" is strictly off-topic unless it helps us understand the WTC7 topic of FFA.

Here is the brief answer:
1) We are discussing the "initiation" stage for Twin Towers. The focus is on the "progression" of failure within the initiation storage. Not the "progression" stage per se;
2) The key mechanism of the "initiation" stage was sequenced cascading failure of columns in axial loading driven by load redistribution;
3) Something must be the first element to start failing ("start" because it is not instantaneous and overlaps with "second", 'third" etc etc);
4) The process is, in principle, the same for spandrel-connected columns as it would be for individual, stand-alone, columns. It just makes the maths several degrees more complicated.
5) "inward bowing" of a section of perimeter spandrel connected columns is identified by NIST as a plausible "first to fail" AKA the "trigger" event to start the cascading failure. I agree it is plausible and highly probable.
6) (This could be a key point for some of your concerns) Where a sector of the perimeter is inward bowed it is weaker than sectors that are lesser bowed or unbowed. The "most bowed" sector will tend to fail first - due to greater eccentricity of loading which means greater reduction of axial compressive strength.

 

[Jeffrey Orling]

Remember that "Twin Towers" is strictly off-topic unless it helps us understand the WTC7 topic of FFA.

Here is the brief answer:
1) We are discussing the "initiation" stage for Twin Towers. The focus is on the "progression" of failure within the initiation storage. Not the "progression" stage per se;
2) The key mechanism of the "initiation" stage was sequenced cascading failure of columns in axial loading driven by load redistribution;
3) Something must be the first element to start failing ("start" because it is not instantaneous and overlaps with "second", 'third" etc etc);
4) The process is, in principle, the same for spandrel-connected columns as it would be for individual, stand-alone, columns. It just makes the maths several degrees more complicated.
5) "inward bowing" of a section of perimeter spandrel connected columns is identified by NIST as a plausible "first to fail" AKA the "trigger" event to start the cascading failure. I agree it is plausible and highly probable.
6) (This could be a key point for some of your concerns) Where a sector of the perimeter is inward bowed it is weaker than sectors that are lesser bowed or unbowed. The "most bowed" sector will tend to fail first - due to greater eccentricity of loading which means greater reduction of axial compressive strength.

Sorry for the derail.
I do subscribe to the notion that the cores of the twin towers grew weaker... ie lost capacity and the mechanism most surely involved load redistribution. It is perhaps more interesting to explore how this process began... For example did heat from wires drive down capacity of a column or columns and the "hand off" of load process began... or was there another mechanism such as loss of bracing driving capacity down leading to bucking.
The hand off (load redistribution) is independent and nothing to do with heat per se.
++++
After plane strike column destruction it seems that heat was what resulted in the load redistribution process. The towers survived the initial plane strikes... which most likely destroyed columns and the loads of those columns were redistributed to remaining columns. Then heat did "something"...
Heat would lower the strength of columns and cause it to expand. Heat would weaken and expand lateral steel.
Did bracing fail in a similar manner as we were told happened in 7WTC around column 79. The steel cross section of column 79 was massive. The columns at the twin tower plane strike zone had much smaller cross sections...and as such more vulnerable to the the effect of heat.
Note that much of the steel bracing were rolled sections on seats that were bolted / welded to the sides of the 3 story columns. It's conceivable that the heated bracing may have pushed the columns causing buckling... or dropped and left increased the unbraced length lowering working strength. Connections might have failed in shear from the expanding beams. Heat did something to cause the columns to continue to weaken, fail and hand off their loads. The process can go runaway... and it did.... core's aggregate capacity was driven down to the point it could no longer support the superimposed loads.

 

[econ41]

Sorry for the derail.

We are going further off this thread's topic - FFA of the WTC7 Facade.

I do subscribe to the notion that the cores of the twin towers grew weaker... ie lost capacity and the mechanism most surely involved load redistribution.

Yes.

It is perhaps more interesting to explore how this process began...

In an appropriate thread. The need in this thread is to recognise and possibly quantify the balance between three probable causes of WTC7 perimeter FFA. viz column buckling, transfer truss failure, direct failure of cantilevers. And almost certain it is not one sole cause but a proportional mix of all three.

For example did heat from [fires] drive down capacity of a column or columns and the "hand off" of load process began... or was there another mechanism such as loss of bracing driving capacity down leading to bucking.

Matters not for most purposes whether direct heat weakening or indirect mechanisms. It is not relevant here in this thread.

The hand off (load redistribution) is independent and nothing to do with heat per se.

Agreed.

++++
After plane strike column destruction it seems that heat was what resulted in the load redistribution process.

Heat or CD and 21 years on nobody has "proved" CD.

The towers survived the initial plane strikes... which most likely destroyed columns and the loads of those columns were redistributed to remaining columns. Then heat did "something"...
Heat would lower the strength of columns and cause it to expand. Heat would weaken and expand lateral steel.
Did bracing fail in a similar manner as we were told happened in 7WTC around column 79. The steel cross section of column 79 was massive. The columns at the twin tower plane strike zone had much smaller cross sections...and as such more vulnerable to the the effect of heat.
Note that much of the steel bracing were rolled sections on seats that were bolted / welded to the sides of the 3 story columns. It's conceivable that the heated bracing may have pushed the columns causing buckling... or dropped and left increased the unbraced length lowering working strength. Connections might have failed in shear from the expanding beams. Heat did something to cause the columns to continue to weaken, fail and hand off their loads. The process can go runaway... and it did...

All truisms but don't progress either the Twin Towers discussion or this current thread discussion.

... core's aggregate capacity was driven down to the point it could no longer support the superimposed loads.

Not just "core". Total capacity was "driven down"

 

[Henkka]

Heat or CD and 21 years on nobody has "proved" CD.

What would you consider proof of CD?

 

[Thomas B]

We are on topic - explaining the necessary foundations of physics needed to answer the OP Topic.

I'm still with you. I'm ready for a model with multiple floors and loads distributed throughout the structure, not concentrated in a "rigid block" or beam on top of the columns, but loading floors that are connected along the length of the columns.

As I understand it, load redistribution in such a structure depends, in part, on whether the connections are pin joints or moment joints. The misunderstanding among truthers, as I understand it (incompletely), is that steel-framed buildings are essentially giant moment-resisting frames.

The perimeters walls, on this view, are sometimes described as Vierendeel trusses. Jim Hoffman quotes a 1985 textbook at his website, but it seems to be something they also get can from Wikipedia:

The buildings used high-strength, load-bearing perimeter steel columns which acted as Vierendeel trusses.[89][85] Although the columns themselves were lightweight, they were spaced closely together, forming a strong, rigid wall structure.[70][90]

https://en.wikipedia.org/wiki/Construction_of_the_World_Trade_Center#Design_elements

Such trusses are made up of rectangular rather than triangular frames, as are common in bridges using pin–joints. Because of the lack of diagonal members, Vierendeel trusses employ moment joints to resist substantial bending forces.

https://en.wikipedia.org/wiki/Vierendeel_bridge

So, like I say, I'm imagining your A-G diagram extended ten stories up, giving us a frame made of 70 60 rectangles. We distribute gravity loads on the floor slabs of each rectangle and specify the initital loads and safety factors on each column section and column-to-floor connection (and we specifiy whether these are pin or moment joints.) We then start severing columns and track the "theoretical" load redistribution, i.e., we draw the load paths around the damage.

Is that about right?

 

[econ41]

What would you consider proof of CD?

Just the standard "proofs" needed for any scientific argument:
1) Falsification of the many extant hypotheses which show that initial damage plus unfought fires were sufficient;
THEN
2) presentation of a viable hypothesis showing how CD could have been achieved.

Those two would establish the prima-facie case for serious discussion.
THEN - proof that CD was actually performed would be the cherry on the top of the cake.

(And there is a bit of overlap in those requirements- It is plausible to prove "CD was performed" without falsifying "CD was not needed - fires were sufficient", All that means is that the brilliant "Inside Jobber" or terrorist who performed the CD wasted their time since it wasn't needed.)

 

[econ41]

I'm still with you. I'm ready for a model with multiple floors and loads distributed throughout the structure, not concentrated in a "rigid block" or beam on top of the columns, but loading floors that are connected along the length of the columns.

I understand that. I'm still biased towards explaining from base principles THEN trying to show how it fits in a much more complicated scenario.

I'm also conscious that the need is to explain why the falling perimeter shell of WTC 7 achieved FFA. And that is a lot easier than a multi-storey complex and generic situation.

As I understand it, load redistribution in such a structure depends, in part, on whether the connections are pin joints or moment joints. The misunderstanding among truthers, as I understand it (incompletely), is that steel-framed buildings are essentially giant moment-resisting frames.

Most truthers don't get to that level of understanding. In fact, neither do most debunkers. The proportion who understand physics at anywhere near the necessary level of complexity is quite small.

The perimeters walls, on this view, are sometimes described as Vierendeel trusses. Jim Hoffman quotes a 1985 textbook at his website, but it seems to be something they also get can from Wikipedia:

The buildings used high-strength, load-bearing perimeter steel columns which acted as Vierendeel trusses.[89][85] Although the columns themselves were lightweight, they were spaced closely together, forming a strong, rigid wall structure.[70][90]

https://en.wikipedia.org/wiki/Construction_of_the_World_Trade_Center#Design_elements

Such trusses are made up of rectangular rather than triangular frames, as are common in bridges using pin–joints. Because of the lack of diagonal members, Vierendeel trusses employ moment joints to resist substantial bending forces.

https://en.wikipedia.org/wiki/Vierendeel_bridge

So, like I say, I'm imagining your A-G diagram extended ten stories up, giving us a frame made of 70 60 rectangles. We distribute gravity loads on the floor slabs of each rectangle and specify the initital loads and safety factors on each column section and column-to-floor connection (and we specifiy whether these are pin or moment joints.) We then start severing columns and track the "theoretical" load redistribution, i.e., we draw the load paths around the damage.

Agreed the summary of engineering issues. BUT I think you are coming at the topic of this thread from the more difficult end of the argument.

The Facade and perimeter shell of WTC 7 fell because support was removed allowing about 7-8 storeys of free fall. What removed the support is the challenge. Not what happened in the falling "block". We know it fell irrespective of its internal mechanics.

Is that about right?

Yes except I think you are chasing the wrong goal. The goal is to explain what removed the support to let the perimeter shell fall with near FFA. The thread OP seems to presume it was buckling. (And "NISTwas right") I don't agree that it has been proven that buckling is the (implied) sole cause.

The internal mechanics of the perimeter shell don't define what removed the support.

 

[Thomas B]

I'm still biased towards explaining from base principles THEN trying to show how it fits in a much more complicated scenario.

... I think you are chasing the wrong goal. The goal is to explain what removed the support to let the perimeter shell fall with near FFA...

The internal mechanics of the perimeter shell don't define what removed the support.

OK. I'll wait until you get to it. I just need to be able to imagine the load paths in the perimeter, both before the support was removed (when the building was standing normally) and after.

After all, the perimeter did not fall over but into itself. It fell (more or less) freely into the plane of its force diagram.

I may not be saying that in a technically correct way; I'm just noting that our ABC model is two-dimensional, so the support has to be removed in the plane, and the collapse must result -- must be theoretically predicted -- along its vertical dimension (which is also where the gravity forces are operating.)

 

[Jeffrey Orling]

The 40 story moment frame and attached curtain wall could only move down as they did as a single unit if

a - the support below which supported them totally and "instantly" lost capacity (to support the moment frame and curtain wall)

b - the support below which supported them totally and "instantly" was displaced (nothing to support the moment frame and curtain wall)

c - the strength of the support of the moment frame gradually "eroded" until the aggregate strength was below that needed to support the moment frame and it dropped... crushing and or displacing the supports as it came down.

graphic depiction of the 4 sides of the building. Black lines represent columns and spandrel steel of the moment frame and trusses in the plane of the facade. Blue represents glass areas in the plane of the facade. Tan is an opening to the basement level.

 

[Jeffrey Orling]

So... My hunch is that the entire interior of the building came down as a result of the collapse of the core and it's perimeter columns (another discussion)... The 47 stories of broken up concrete slabs, contents, mechanical equipment and steel beams and columns tried to "squeeze" itself into the shell of the building at its base. The chaotic fall mass could not be contained and the axial structure at the base, bulged and buckled or was pushed outward... or all three resulting in the moment frame and attached curtain wall losing axial support. Without support it collapsed.
Photos show the debris as tall as 8 or 9 stories and extending over the sidewalks and streets surrounding the building.

 

[Jeffrey Orling]

This is a cartoon of the 47 floors of debris at the base showing the forces pushing outward at the support of the moment frame and curtain wall:

 

[Henkka]

Just the standard "proofs" needed for any scientific argument:
1) Falsification of the many extant hypotheses which show that initial damage plus unfought fires were sufficient;
THEN
2) presentation of a viable hypothesis showing how CD could have been achieved.

Those two would establish the prima-facie case for serious discussion.
THEN - proof that CD was actually performed would be the cherry on the top of the cake.

But what could "proof that CD was actually performed" be in your opinion? And if someone could present that proof, wouldn't steps 1) and 2) be rather redundant? If there's proof of CD, it wouldn't matter if the truther could falsify every possible fire-induced collapse hypothesis, or present a "viable" explanation of how exactly it was done.

 

[Landru]

But what could "proof that CD was actually performed" be in your opinion? And if someone could present that proof, wouldn't steps 1) and 2) be rather redundant? If there's proof of CD, it wouldn't matter if the truther could falsify every possible fire-induced collapse hypothesis, or present a "viable" explanation of how exactly it was done.

This is off topic.

 

[econ41]

But what could "proof that CD was actually performed" be in your opinion? And if someone could present that proof, wouldn't steps 1) and 2) be rather redundant? If there's proof of CD, it wouldn't matter if the truther could falsify every possible fire-induced collapse hypothesis, or present a "viable" explanation of how exactly it was done.

Please read ALL of what I said - and if you are still unclear - take the question to an appropriate thread. I will follow @Landru's guidance as to "off-topic".

 

[econ41]

The 40 story moment frame and attached curtain wall could only move down as they did as a single unit if

a - the support below which supported them totally and "instantly" lost capacity (to support the moment frame and curtain wall) OR

b - the support below which supported them totally and "instantly" was displaced (nothing to support the moment frame and curtain wall) OR

c - the strength of the support of the moment frame gradually "eroded" until the aggregate strength was below that needed to support the moment frame and it dropped... crushing and or displacing the supports as it came down.

A neat summary of what should be the starting point assumptions. I've made the implied logical OR explicit.

So what caused the loss of support?
(a) Columns buckling;
(b) Cantilevers failing; OR
(c) Transfer truss failures?

My bet is still on a mix of all three in some sort of combination.

And we cannot legitimately claim any single one or any pair as the cause until we eliminate the others.

 

[econ41]

This is a cartoon of the 47 floors of debris at the base showing the forces pushing outward at the support of the moment frame and curtain wall:

Your cartoon is a valuable starting point for the next stage of discussion.

And I agree with your suggestion that core debris accumulating at the lower levels was the plausible cause of the failures we are discussing.

So I'll "borrow" your diagram, @Jeffrey Orling, as a summary of where we should all be agreeing on the macro mechanism of the collapse:

Do we agree on these points:
1) Support failed at the lower levels?
2) The core had mostly collapsed - had no significant part to play at this time?
3)The perimeter shell is mostly intact with some possible flexures that had no significant effect on the observed FFA?

This therefore suggests:
4) that we don't need to consider the issues of flexure or failure in those higher levels.

Which disagrees with your suggestions @Thomas B >> So we may have a need for more discussion of your suggested way forward. I'm saying we don't need to consider what happened above the red dotted line. We should agree that the failures causing FFA were below that line. (Which doesn't preclude that we may need to define exactly where the line is. I haven't made the location specific --- yet!)

Thoughts? Are we ready to consider the balance of contributions to causing the observed FFA OR do we accept the implications of the OP that "buckling" was the sole cause?

 

[Thomas B]

Which disagrees with your suggestions @Thomas B >> So we may have a need for more discussion of your suggested way forward. I'm saying we don't need to consider what happened above the red dotted line. We should agree that the failures causing FFA were below that line.

I think we may be talking past each other. You've skipped ahead to "what actually happened", but we were working on a "theoretical prediction" from a simple 7-column diagram. As long as we can at some point bridge that gap, all is well.

We need a simple load path diagram in which a progressive series of failures cause a "red line" (across all the columns at the same level) to move downwards, while remaining horizontal, so that the roofline descends (in essentially free fall). The failures have to begin with the failure of a single (or a few) columns or floor connections.

 

[econ41]

I think we may be talking past each other. You've skipped ahead to "what actually happened", but we were working on a "theoretical prediction" from a simple 7-column diagram.

"Explaining a principle" which is one of the foundation issues of a valid response to the OP.

Yes - I both "skipped ahead" AND challenged members to agree where we should be at - with one exception in your case. Setting an "aiming mark" for where we need to GOP and creating a gap to be bridged.

So we know there is a gap to bridge.

As long as we can at some point bridge that gap, all is well.

Until we agree on the "start line" we can not bridge the "gap" because members will be thinking of different gaps.

So:
Do you agree on these points:
1) Support failed at the lower levels?
2) The core had mostly collapsed - had no significant part to play at this time?
3)The perimeter shell is mostly intact with some possible flexures that had no significant effect on the observed FFA?

...and this one where you and I were down different paths:
4) that we don't need to consider the issues of flexure or failure in those higher levels.

If we agree - we can move on. If "we" don't agree on the start line we need to resolve any differences. (And I think "we" need to include @Henkka and @Jeffrey Orling or they may go down divergent paths.)

We need a simple load path diagram....

Possibly - I'm not sure of what you envision for the representation. And we may have one more step to take before we get to that.

in which a progressive series of failures cause a "red line" (across all the columns at the same level) to move downwards,

Yes. A minor proviso - the elastic flexing of that "upper section" above the line will (may??) (could??) mean that the red line is not dead straight. We can deal with that aspect if it becomes critical to our discussion.

while remaining horizontal, so that the roofline descends (in essentially free fall).

Yes.

The failures have to begin with the failure of a single (or a few) columns or floor connections.

By definition - there must be a sequenced cascading process of load re-distribution in a "fire caused" event. Not so for CD but we are not considering CD at this time.

And, although we will probably never know which of the three possibilities was the trigger, it must have been the direct trigger of either columns buckling or cantilever failure or the indirect result of transfer truss failure. (And there could be some overlapping if and when we need to explore the details.)

So can we (you and I at this time) agree that the single or few elements trigger you identify ( "have to begin with the failure of a single (or a few)...") has to be ONE of the three possibles? And that, at this stage of the debate, we cannot say which OR eliminate any of the three?

 

[Thomas B]

in which a progressive series of failures cause a "red line" (across all the columns at the same level) to move downwards, while remaining horizontal, so that the roofline descends (in essentially free fall).

Yes. A minor proviso - the elastic flexing of that "upper section" above the line will (may??) (could??) mean that the red line is not dead straight. We can deal with that aspect if it becomes critical to our discussion

It occurs to me that it might be easier just to think of the "red line" as the roofline. We need a diagram of a structure in which local damage to a few columns and connections near the bottom cause the free fall of the roofline. It's not important to identify the region of the building in which

we don't need to consider the issues of flexure or failure

from the outset. It will emerge as a result of the model that theoretically predicts a level roofline falling freely.

 

[Thomas B]

challenged members to agree

Until we agree

Do you agree

If we agree - we can move on. If "we" don't agree on the start line we need to resolve any differences.

So can we (you and I at this time) agree

I have to admit I find it hard to keep track of all these agreements we have to make. So I apologize in advance for not reacting to each one. I seems like we're more or less on the same page, though.

I'm not sure of what you envision for the representation.

I tried to explain it in my post #376.

I'm imagining your A-G diagram extended ten stories up, giving us a frame made of 60 rectangles. We distribute gravity loads on the floor slabs of each rectangle and specify the initital loads and safety factors on each column section and column-to-floor connection (and we specifiy whether these are pin or moment joints.) We then start severing columns and track the "theoretical" load redistribution, i.e., we draw the load paths around the damage.

What is it you find unclear about this vision?

 

[Jeffrey Orling]

So the problem may be how explain the collapse of the floors inside the perimeter of the building.

The floors were supported at the perimeter on pinned connections.
The floors on the core side were supported by girders around the core's perimeter.

Fire on the floors 12-14 around columns 69, 80 and 81 led to buckling when the heat expanded beams and girders slid off / were pushed off the beam seats at those columns.

EPH lost support and dropped through into the roof and probably downward through the building. There is evidence of this in distortion of the curtain wall glass.

Columns which have lost their bracing are weaker and prone to buckling.

The floor mass of the stories above 12-14 on the east side crashed down on to the transfer structure on floors 5-7 dislodging the transfers.

TT1 and TT2 as a result likely disturbed the line of girders on the north side of the core and "upset" the MG27 girders supporting the floors north of the core and undermining the north side of the moment frame.
cz

The diagram shows how the West, North and East side of the core's perimeter at floors 5-7 were disturbed. The collapse / movement of the girders from col 75-to col 81 likely disturbed / displaced / dislodged the girder on the all sides of the core ... east=west magenta girders. If this disruption of the girders around the core occurred.... and it's a good possibility that they were disturbed.... All core side floor support was destroyed leading to the collapse of the floors system of the entire footprint. The was likely at or below from 12-14 but not lower that floors 5-7.
The building was hollowed out and as the previous cartoon shows the debris pile likely led the buckling of the 27 axial support location from the ground up to floor 8. This distance accounts for the 2.25 second near FFA of the moment frame and curtain wall.

 

[econ41]

I have to admit I find it hard to keep track of all these agreements we have to make.

I have reduced the explanation I am presenting to a series of separate and simple steps, A step by step progression of reasoned steps. NOT "one big confused mess". That process should be fundamental to any reasoned explanation.

So I apologize in advance for not reacting to each one. I seems like we're more or less on the same page, though.

We will be "on the same page" IF:
(a) You agree with each simple step I have isolated, and in the sequence I present them; OR
(b) You identify any step you disagree with and tell me why you disagree so that we can discuss it further. Until we agree there is no point in trying to move forward whilst we are not clear on what we agree.

I tried to explain it in my post #376.

What is it you find unclear about this vision?

What is unclear is why you see a need to explain the unexplainable complicated bit. That is why I drew the "red line" across the structure and gave my reasons. We don't need to understand the details "above the line" OTHER than the perimeter shell fell as an integral whole structure:

,,"Above the Red Line" is the perimeter shell structure that fell integrally with some minor flexures.
"Below the Red Line" is where the support was lost.
Our discussion is about what caused the loss of support. Buckling, transfer truss failure or cantilever failure.

Above the red line where you want to go has no significance for the question we are addressing AND we will never be able to analyse that "Top Shell" in the manner you propose.

I'm saying we should focus on what we probably can handle, that is the topic of this thread and not be confused by irrelevancies.

AND I'm presenting a reasoned step-by-step argument to support my claim.

 

[econ41]

So the problem may be how explain the collapse of the floors inside the perimeter of the building.

That is not the current problem. The FACT of core-collapse should be an agreed premise for the start of current discussions

The diagram shows how the West, North and East side of the core's perimeter at floors 5-7 were disturbed. The collapse / movement of the girders from col 75-to col 81 likely disturbed / displaced / dislodged the girder on the all sides of the core ... east=west magenta girders. If this disruption of the girders around the core occurred.... and it's a good possibility that they were disturbed.... All core side floor support was destroyed leading to the collapse of the floors system of the entire footprint. The was likely at or below from 12-14 but not lower that floors 5-7.

Yes.

The building was hollowed out and as the previous cartoon shows the debris pile likely led the buckling of the 27 axial support location from the ground up to floor 8.

Yes

This distance accounts for the 2.25 second near FFA of the moment frame and curtain wall.

Yes.

Now all of that supports transfer truss and/or cantilever failure. And both of those together with column buckling are the causes of FFA we are supposed to be discussing. Except @Thomas B is unclear of the process of discussion.

 

[Thomas B]

What is unclear is why you see a need to explain the unexplainable complicated bit.

I'm not sure what bit you mean.

 

[Thomas B]

Our discussion is about what caused the loss of support.

Maybe we mean the same thing, but I thought the discussion was about how the progressive loss support under the facade led to its free fall.

 

[econ41]

Maybe we mean the same thing, but I thought the discussion was about how the progressive loss support under the facade led to its free fall.

Yes. And, subject to your agreement being achieved.

I have isolated:
(a) That the loss of support occurred below the Red Line and
(b) involved three (3) factors, two of which are partially overlapping.

AND prima facie none of the micro-detail of the Top Shell - above the red line - had a significant impact.

BECAUSE:
(i) Column buckling results from overload independent of the micro-detail of how the overload was applied;
(ii) Cantilever failure simply removes the support independent of what caused the failure;
(iii) Transfer Truss failure has the same effect except it operates indirectly in removing support.

AND none of that requires an understanding of the stress and strain effects in the flexures of the "Top Shell" above the red line. You are making explanation impossible for no valid reason.

 

[econ41]

I'm not sure what bit you mean.

The "Top Shell" above the Red Line. We can treat it as a "Black Box" - we don't need to understand internal workings at this stage of discussion if ever.

I drew a red line on the diagram:
(a) the bit under the red line is where the loss of support occurred and we can discuss what caused the loss of support in that area. << THAT is the topic of this thread.

(b) The perimeter shell structure above the Red Line is the bit - the "part" - the "portion of the structure" - that we don't need to discuss. BUT you persist in wanting to consider it. I've laid out and repeated my reasoning why we don't need to go there.


We are essentially stalled in discussion.

 

[Jeffrey Orling]

As the shell above the red line moved as a single "thing"... without floors attached... or it was the entire building above the red line which came down as a whole.

However there is visual evidence that there was interior "collapse" on the East side below the EPH... and it appears that what fell went down right to the ground.

The collapse of the structure and the floors on the east side was a suspect for disruption of the transfer... below the red line on floors 5-7. This is a suspect for undermining much of the interior... which included the floors from 8 to the roof.... essentially leaving a hollow shell behind "trying" to contain all the floor debris inside the shell at the base. It likely could not and as the cartoon above shows... that debris undermined the axial support for the moment frame up to floors 5-7 where the was a perimeter belt truss under the moment frame.

It's unlikely that the moment frame experience new loads. as the floor collapse would "unload" it

 

[econ41]

As the shell above the red line moved as a single "thing"...

Agreed - it is the starting premise.

without floors attached...

Without any significant effects from remnants of floors which could still be attached.

or it was the entire building above the red line which came down as a whole.

Why do you regress? Going back to question decisions already made in the presented argument?

However there is visual evidence that there was interior "collapse" on the East side below the EPH... and it appears that what fell went down right to the ground.

The collapse of the structure and the floors on the east side was a suspect for disruption of the transfer... below the red line on floors 5-7. This is a suspect for undermining much of the interior... which included the floors from 8 to the roof.... essentially leaving a hollow shell behind "trying" to contain all the floor debris inside the shell at the base.

Already agreed. Why circle back?: Except for the orange part which is plausible at this stage and may become significant if the discussion progresses.

It likely could not and as the cartoon above shows... that debris undermined the axial support for the moment frame up to floors 5-7 where the was a perimeter belt truss under the moment frame.

Plausible but it is an aspect for further discussion if it becomes significant to the argumentation.

It's unlikely that the moment frame experience new loads. as the floor collapse would "unload" it

Plausible but no immediate relevance - it is "above the Red Line".