'What about building 7'?

[gerrycan]

Being no more than an intern in this discussion, watching the pros battle it out, I may be out on a limb now, but I'll say it anyway:

This is getting most silly!

On several floors, many beams, girders, columns, concrete slabs and what else have your were subject to a history of heating and cooling caused by wandering fires, right?
So hundreds of elements in that 16-story assembly that NIST considered were subject to all sorts of heat-induced contortions - expanding, contracting, bending, torquing, ... right?
Each node thus was likely to move some distances up, down, N, E, W, S by different distances at different times - right?
And those distances would not only depend on the heat-induced deformation of the elements that come together at that node, but also on the displacememts of all the neighboring nodes, which in turn are affected by displacements of nodes that they link to - right?
NIST ran an FEA with ANSYS, which resulted in such displacements of nodes - right?
Their model produced as implicit and explicit results displacements of every node at many points in time, right?
And quite generally, we do not know the values of almost all such displacements, right?
We know them neither in the NIST model, nor in the reality of the burning building, right?

So how can you pick out one single element, claim a single value of displacement or non-displacement or whatever that 1 inch is that you are talking about, and try to make us believe that somehow this particular inch makes or breaks the overall outcome of "fire induced floor failure at col. 79"? This appears to me to be exceedingly silly!

K3004 is the longest of the floor beams to the East of the C79-44 girder.
At the opposite end to the one framing into the girder it connects to C38, which in NISTs analysis is taken to be infinitely strong. Given that the girder connection is said to fail due to expansion, is it not reasonable to think that the opposite connection would fail similarly? I know the outside of the building will be cooler, but steel does conduct. This failure at C38 would mean that the beam would push 1" less to the West.
If we are going to allow for column shift to the east this should be quantified and demonstrated possible at least, as has the failure of the C38 connection.

 

[DGM]

K3004 is the longest of the floor beams to the East of the C79-44 girder.
At the opposite end to the one framing into the girder it connects to C38, which in NISTs analysis is taken to be infinitely strong. Given that the girder connection is said to fail due to expansion, is it not reasonable to think that the opposite connection would fail similarly? I know the outside of the building will be cooler, but steel does conduct. This failure at C38 would mean that the beam would push 1" less to the West.
If we are going to allow for column shift to the east this should be quantified and demonstrated possible at least, as has the failure of the C38 connection.

You didn't answer his questions.

 

[Spanx]

I still can't work out why Gerrycan is not arguing with the NIST and has chosen to argue here instead

Hang yes I can work it out.

 

[Oystein]

K3004 is the longest of the floor beams to the East of the C79-44 girder.
At the opposite end to the one framing into the girder it connects to C38, which in NISTs analysis is taken to be infinitely strong. Given that the girder connection is said to fail due to expansion, is it not reasonable to think that the opposite connection would fail similarly? I know the outside of the building will be cooler, but steel does conduct. This failure at C38 would mean that the beam would push 1" less to the West.
If we are going to allow for column shift to the east this should be quantified and demonstrated possible at least, as has the failure of the C38 connection.

I don't get exactly what problem you're claiming - do you wish NIST had provided a couple more details than they did? Or do you claim that the NIST model failed to predict some ocurrance that you "think" would happen (according to what - your imagination? engineering judgement? calculations?)?

But you fail to address 90% or more of my post. I must be the 20th person at least who tries to explain to you that the entire assembly experienced deformations and displacements in all directions, the magnitudes of which we generally do not know because NIST did not publish all that data, and that these unknown displacements make it impossible for us to know exactly which element must have expanded/moved/bent by how much to reach some threshold. It follows that the analysis can't hinge on 1 inch here or there.

Having said this, I concede, even without argument, that of course the result of NIST's simulation could be wrong, even to the extent that said girder did not walk off where, when and how NIST says it did.
In fact, I will state as my firm belief that the result of NIST's simulation is definitely "wrong" in thousands upon thousands of details. Perhaps they did not get a single number right - that would largely depend on the level of accuracy you demand.
And yet the simulation is probably valid. At least your approach - questioning single numbers - is not a valid way to invalidate the results.

NIST shows that the assembly will experience events that are capable of bringing the structure close to collapse in a manner that is consistent with observations. Your attempt to argue that these events fall 1 or 2 or 3 inches short of actually initiating the collapse are refuted by the simple reality that evidently the collapse DID initiate, with columns 79 and 80 being the first columns to go. Parsimony demands that even IF the model falls short by 1 or 3 inches, that is due to mere inaccuracy of the model (and/or inaccuracy of the manufacturers and assemblers of the steel elements in matching the structural drawings and specifications).

 

[Reactor drone]

Fair enough. But we should also therefor include the failure of K3004 connection at C38. There goes another inch right there.

The trouble is that the FEA shows only partial damage to that connection in case C and none at all in case B.

 

[chrismohr]

Thanks for the breath of fresh air Oystein.

 

[gerrycan]

The trouble is that the FEA shows only partial damage to that connection in case C and none at all in case B.

Yes. That's the trouble right there. These were taken to be infinitely strong in the model.
In reality k3004 would expand and break this connection at C38 leaving NIST about another inch to find. Keep in mind that expansion is assumed to be only in the direction that favours NISTs hypothesis up until now.

 

[DGM]

Yes. That's the trouble right there. These were taken to be infinitely strong in the model.
In reality k3004 would expand and break this connection at C38 leaving NIST about another inch to find. Keep in mind that expansion is assumed to be only in the direction that favours NISTs hypothesis up until now.

By "infinity strong", that means there was no movement in reality? Why are you willing to accept this assumption and not others?

Why are you so focused on a single area? Was there no big picture?

BTW: Are you winning?

 

[LSSBB]

Yes. That's the trouble right there. These were taken to be infinitely strong in the model.
In reality k3004 would expand and break this connection at C38 leaving NIST about another inch to find. Keep in mind that expansion is assumed to be only in the direction that favours NISTs hypothesis up until now.

Now take all the possible expansions and distortions and all the possible offsets. What is the probability distribution of possible position outcomes looking like?

 

[DGM]

Now take all the possible expansions and distortions and all the possible offsets. What is the probability distribution of possible position outcomes looking like?

I think you would need a large amount of computing power to answer this............

 

[beachnut]

Now take all the possible expansions and distortions and all the possible offsets. What is the probability distribution of possible position outcomes looking like?

On 911 it looks like a building, failed in fires not fought. Ironically a building that was totaled by fire before it fell; if it had stood, it was still totaled by fire. I can't believe he can't explain, or support his theory.

 

[LSSBB]

I think you would need a large amount of computing power to answer this............

What, you can't just muddle through it with hand calculations, backs of envelopes and posting on obscure forums??

The problem gerrcan is running into is that the probability cloud is currently including NIST being right, and the building collapsing due to fire.

Note also that NIST being right is a subset of the "building collapse due to fire" group - it is possible for NIST to be wrong on details and the building still having collapsed due to fire.

 

[Reactor drone]

Yes. That's the trouble right there. These were taken to be infinitely strong in the model.
In reality k3004 would expand and break this connection at C38 leaving NIST about another inch to find. Keep in mind that expansion is assumed to be only in the direction that favours NISTs hypothesis up until now.

If the connections were infinitely strong how did they get damaged in the model? As far as I know the columns were only fixed at the top and bottom in the 16 storey model so the possible bulging of the exterior was modelled and they still got expansion required in the model to meet their lateral failure criteria.

 

[gerrycan]

By "infinity strong", that means there was no movement in reality? Why are you willing to accept this assumption and not others?

Why are you so focused on a single area? Was there no big picture?

BTW: Are you winning?

It's an assumption that should have been accounted for when NIST were setting the point at which they could deem the girder to have failed, initiating the collapse progression in their hypothesis.
The reason for focus on the NE of the building is because NIST say this is where collapse initiated.

 

[DGM]

It's an assumption that should have been accounted for when NIST were setting the point at which they could deem the girder to have failed, initiating the collapse progression in their hypothesis.
The reason for focus on the NE of the building is because NIST say this is where collapse initiated.

They said a probable initiation point. Was this expansion the only factor they used to determine this? You seem to imply it is.

BTW: Why did you shy away from backing up your claim about code recommendations?

 

[Animal]

So after posting that you get this reply:



This is similar to the previous discussion on solution convergence.

The real issue here is that two groups of people are having two very different conversations.

Group 1:
The group supporting an inside job lacks the foundational/fundamental knowledge that comes from education + experience and as a result of that their specific technical understanding lacks depth and their general technical knowledge is not very broad.

Thus in their mind the discussions over the methodologies that NIST used are actually DEBATES where they are making valid points and then the "debunkers" are making counter points.

To them this thread is actually a real debate where they can "win" or score points by finding supposed errors in the NIST analysis.

Group 2:

The debunker side, particularly those who are highly technical, are attempting to EDUCATE the truthers as to why they are wrong and how engineering analysis works in the real world. There is no actual "debate" as to whether or not they are wrong; instead there are efforts by several posters to explain these concepts to them.

After all of this discussion I doubt even LSSBBs explanation on solution convergence is understood, much less the more involved topics.

As long as group 1 believes this is an actual debate the attempts by group 2 to educate, correct, and guide group 1 to a better understanding will only serve to reinforce group 1's delusion about what this discussion really is.

There is a reason why the vast majority of the scientific and engineering world simply ignore them.

Just like the troofer handwave away of the possibility of leveraging......even if the longest beams were still connected it does not mean they would necessarily constrain the leveraging.......troofers cannot think in complex scenarios.....only paper napkin ones.

 

[Animal]

Now take all the possible expansions and distortions and all the possible offsets. What is the probability distribution of possible position outcomes looking like?

And just because the fire simulation signified certain things happening does not mean that they happened in the exact sequence, or at the exact temps etc.etc.

Troofers just seem to want to stay on the hamster wheel why dicky gage travels at their expense.

 

[LSSBB]

BTW: Why did you shy away from backing up your claim about code recommendations?

Maybe it was the Gage quote from Chris.

 

[Tony Szamboti]

I am illustrating that Grizzly took the CTE of steel for 3 separate temperatures over a range and then reapplied those 3 coefficients separately over the entire range each time.
He should have taken those 3 values and applied them to the temperature curve and arrived at one figure for expansion, not 3.
His post gave the impression that 6.37" of linear expansion is possible in a 53ft beam at 600C from room temp, and it just isn't.

They also need to be told that the steel for the beams and girder was ASTM A572, in case they actually want to find the average CTE from room temp to 600 degrees C for it, so they can see for themselves that the 5.5 inch maximum beam expansion you have told them is right.

It was interesting to see that some who seem to doubt what you are saying didn't know the steel specification.

 

[Grizzly Bear]

I am illustrating that Grizzly took the CTE of steel for 3 separate temperatures over a range...

Same temperature, 3 steel alloys (ergo 3 coefficients). Thus the low, mid, and high range values. I appreciate critique, but it again needs to be based on understanding what I did, and how far I applied it. Even after explaining this, it's apparent you're not following.

His post gave the impression that 6.37" of linear expansion is possible in a 53ft beam at 600C from room temp, and it just isn't.

It is possible in that range for certain steel alloys, that's the only point my exercise set out to demonstrate. I used three alloys to approximate the most likely type of steel/metal to be used in a similar calculation. To calculate the final value and it's effect - which includes the structural response of adjacent members - requires more than my simplified calculations, and for that matter your apparent critique of them.

This is why, these little meta detail discussions go nowhere. You're fixated on one thing, when it's been conveyed to you in no unclear terms that there are other factors at play that will affect that real outcome. If you intend to continue misreading or misrepresent my methodology, you do so at your own peril.

But again, this is one of those discussions asking which straw broke the camels' back. Not a question of whether or not the fires caused the collapses. The general consensus is that they did, but the specific failure under such conditions is essentially what you're stuck dealing with.