Why did NIST ignore thermal failure of the core beams of the Twin Towers?

Abdullah

Active Member
The coefficient of linear thermal expansion of steel is around 1/100,000 per °C.

This means that at a temperature of 420°C, a 5m steel beam will:

1. Attain 800MPa elastic stress under restrained expansion

2. Expand 2cm under unrestrained expansion.

The Twin Towers beams were made of steel with a yield stress of 248MPa or 345MPa. This means that in case of #1, the beams would have buckled. This buckling could lead to beam collapse. What effect would that have?

1. It could pull in the exterior walls

2. It could cause core column buckling

The core columns in the impact zones were mainly 36cm and 30cm wide flange sections of 248 or 290MPa, with a safety factor of two. Euler buckling theory:

Length² × boundary condition² × stress = π² × gyradius² × elastic modulus

This gives a buckling length of around 40m, or 11 stories, for the core columns under normal loads and without accounting for stiffness reduction by splices.

In NIST NCSTAR 1-6C we readScreenshot_2022-09-21-17-02-48-818_com.microsoft.office.word-01.jpeg
Core beams were modelled as rigid beam elements. Meaning they can't deform or fail. Basically, NIST has decided not to analyze them as part of the full floor model. Why?

Screenshot_2022-09-21-17-05-12-345_com.microsoft.office.word-01.jpeg
Notice how there is no mention of core beam snslyses. No answers here either. Why?

Here is what NIST says in NCSTAR 1-6 about how the core was weakened and how the floors contributed to collapseScreenshot_2022-09-21-18-53-26-992_com.microsoft.office.word-01.jpeg
Can we really accept

1. NIST's implicit conclusion that core weakening was not caused by floor collapse and

2. Their more explicit conclusion that inward bowing was caused by truss sagging alone?
 
Last edited:

econ41

Senior Member
Notice how there is no mention of core beam snslyses. No answers here either. Why?
Why is it needed? If not needed it is irrelevant.
Can we really accept
Why do "we" (or "you") need to accept? Does your interpretation of either of these two aspects disagree with your own understanding of the collapses? If so - post your explanation for discussion.
1. NIST's implicit conclusion that core weakening was not caused by floor collapse and.....
How does either factor change NIST's conclusion? If in fact what you interpret as an "implicit conclusion" is what NIST somehow relied on?
2. Their more explicit conclusion that inward bowing was caused by truss sagging alone?
Inward bowing was NOT "caused by truss sagging ALONE". Yes IB was near certainly triggered by truss sagging. BUT sagging only needed to start column misalignment with the resulting weakening due to eccentric loading. Once the critical point was reached the inward bowing would self-sustain without further help from joist sagging.
 

Abdullah

Active Member
Why is it needed?
Because the failure of core beams could have brought down the towers, As I said
1. It could pull in the exterior walls

2. It could cause core column buckling

Why do "we" (or "you") need to accept?
Because then we can use that knowledge
- to debunk conspiracy theories
- to design safe buildings

Yes IB was near certainly triggered by truss sagging. BUT sagging only needed to start column misalignment
My point is core beam collapse could have "triggered" IB.

How does either factor change NIST's conclusion?
It could mean that thermal weakening was not the cause of core column failure. Which in turn could invalidate NIST's conclusion that the core columns would not have weakened without insulation damage.
Screenshot_2022-09-22-07-33-59-408_com.microsoft.office.word-01.jpeg
 
Last edited:

johnny plectrum

New Member
I think there’s a lot of ‘ifs’ in there.
Depending on the particular 5m beam, its connectivity to the slab over, its temperature change and its overall restraint conditions, yes, it will expand and it will generate expansion forces.
If those forces were large enough to cause the beam to buckle then there may be floor sag, catenary action and even collapse.
Equally, the loads will simply redistribute to stiffer members including those, on occasion, previously non-loadbearing.

What effect would that have?

1. It could pull in the exterior walls

I struggle to imagine a scenario where tension is induced into the core floors enough to load the perimeter core columns enough to load the office floorplates enough to pull in, or even contribute to the pulling in of the perimeter columns.

2. It could cause core column buckling

Indeed. Loss of areas of the core flooring system could alter restraint conditions for core columns.

Core beams were modelled as rigid beam elements. Meaning they can't deform or fail. Basically, NIST has decided not to analyze them as part of the full floor model. Why?

I believe clause 7 tries, but possibly fails, to clarify how they simplified their modelling of core beams off-grid and with moments due to eccentricity, by introducing additional rigid elements.
It doesn't imply that the core beams were modelled as rigid beam elements.

Can we really accept

1. NIST's implicit conclusion that core weakening was not caused by floor collapse and

I don't see that implied.

2. Their more explicit conclusion that inward bowing was caused by truss sagging alone?

Again, I don't think they have made that conclusion have they?

I don't believe you've made a case that NIST ignored thermal failure of the core beams.
They modelled the core beams within their analysis but, maybe due to its beam and slab simplicity, they didn't dwell on it - unlike the trussed areas?
 

Abdullah

Active Member
I struggle to imagine a scenario where tension is induced into the core floors enough to load the perimeter core columns enough to load the office floorplates enough to pull in, or even contribute to the pulling in of the perimeter columns.
If the edge girder collapsed near the tower corners, it could pull walls in.
I don't see that implied.
It's implied by their not dwelling on it.
I believe clause 7 tries, but possibly fails, to clarify how they simplified their modelling of core beams off-grid and with moments due to eccentricity, by introducing additional rigid elements.
It doesn't imply that the core beams were modelled as rigid beam elements.
I am at a loss as to how U missed ",connected to slab by rigid beam elements". You're right.


That debunks the OP - up to a point. "Not dwelling on it" is still ignoring it. They simply do not examine beam failure. And thatsvs problem.
 

Latest posts

Top