WTC7 and the girder walk-off between column 79 and 44

[Tony Szamboti]

Again...we are talking design loading vs. maximum capacity.

And yes...I looked at the link.

Are you trying to say they would fail with a smaller load?

 

[MileHighMadness]

BTW...your girder doesn't have to walk-off the column seat, for the column to fail in buckling. Once the 4 - 7/8" bolts shear off, the column loses lateral support from the girder. The effective column length is greatly increased, the critical buckling load is greatly reduced and the column fails.

 

[Tony Szamboti]

BTW...your girder doesn't have to walk-off the column seat, for the column to fail in buckling. Once the 4 - 7/8" bolts shear off, the column loses lateral support from the girder. The effective column length is greatly increased, the critical buckling load is greatly reduced and the column fails.

The column would not lose support from the girder when it is pressed up against it due to thermal expansion. Did you forget about that?

The force generated by thermal expansion is much greater than that provided by the four bolts and that is why the girder sheared those bolts. You agreed with that and you should have.

Besides there was no fire at the 10th floor so if the girder doesn't fall it can't take out that girder. With the load on it and its moment of inertia, Column 79 would need to be unsupported laterally for more than five stories before buckling.

 

[MileHighMadness]

Are you trying to say they would fail with a smaller load?

No...I'm telling you there is a difference being what load is design for...and the actual capacity of a structural system is.

 

[MileHighMadness]

The column would not lose support from the girder when it is pressed up against it due to thermal expansion. Did you forget about that?

A column will buckle in either direction...if the girder is pressed up against the column, then it will help the column buckle at a lower axial load.

 

[Tony Szamboti]

No...I'm telling you there is a difference being what load is design for...and the actual capacity of a structural system is.

You are talking in circles, and I suspect it is because there is only one way you want to answer, and you have no basis for it.

 

[Tony Szamboti]

A column will buckle in either direction...if the girder is pressed up against the column, then it will help the column buckle at a lower axial load.

None of the girders on the 6th, 10th, and 14th floors were disconnected from the column so it was supported in both lateral directions with only four stories possibly unsupported. Conservative analyses based on the AISC criteria show the column could go a significantly greater distance than that without lateral support.

Your answers also imply that you realize, after seeing what I sent you, that the axial walk-off postulated by BasQue Arch was also impossible.

 

[MileHighMadness]

You are talking in circles, and I suspect it is because there is only one way you want to answer, and you have no basis for it.

You are starting to sound as obtuse as some of the others here who can't get it through their heads that the NIST girder walk-off failure scenario is impossible, which is the actual premise of this thread.

Tony...I sorry if you think I'm sounding obtuse, but this is basic structural engineering. Once the column/girder connection fails, the column loses critical lateral support and fails in buckling. Walk-off isn't important...IMHO

Later...

 

[Tony Szamboti]

Tony...I sorry if you think I'm sounding obtuse, but this is basic structural engineering. Once the column/girder connection fails, the column loses critical lateral support and fails in buckling. Walk-off isn't important...IMHO

Later...

Please perform an analysis and send it to me. It doesn't have to be elaborate.

 

[MileHighMadness]

Please perform an analysis and send it to me. It doesn't have to be elaborate.

Will do...it will probably be later in the week, I'm travelling M/W.

 

[enik]

Will you post it here so we can all see the results?

 

[erwinl]

Actually you are not correct here. Velocity from distance measurements over time is not determined based on one data point and its difference from the point before. It requires a differencing algorithm which needs the difference between the points before and after a specific point and the time elapsed between the before and after points. That is what we did in the Missing Jolt paper and it shows no velocity loss or deceleration. This was the same methodology as that used by NIST and David Chandler in their measurements of WTC 7. Chandler also measured the North Tower and found no deceleration.

In their WTC 7 measurements both Chandler and NIST had a data point here or there in the overall data, which if measured strictly from the preceding point would indicate greater than or less than freefall during that time. Do you propose that the NIST and Chandler measurements of WTC 7 do not show it was in freefall for the 2.25 seconds they show it was? Or that it was speeding up or slowing down during that time frame?

I'm confused.
Are you suggesting that something, which is in freefall, doesn't speed up?
It does look like you're mixing accelleration and speed like in this post (3450).

 

[Oystein]

I was talking about static failure loads in the post you initially responded to. You don't seem to want to admit that the NIST FAQ proved I was right. The floors could take a static load of 12 floors before failing. This is why I say there would have had to be about eight to nine stories of rubble before quasi-static failures could take place with the floors.

Your source correctly states that this static load of 12 floors is equivalent to a load of 6 floors being applied suddenly. Can you explain briefly what they mean by "suddenly"? Is that after dropping from a certain height? What drop distance would that be?

(Hint: I know the answer, just want to see if you know it, too)

 

[Oystein]

...That is what we did in the Missing Jolt paper and it shows no velocity loss or deceleration.

Do you propose that the NIST and Chandler measurements of WTC 7 do not show it was in freefall for the 2.25 seconds they show it was? Or that it was speeding up or slowing down during that time frame?

I do propose that: The NIST and Chandler (and the more accurate femr2) measurements of WTC 7 do not show it was in freefall for the 2.25 seconds.

Instead, they show that a part of the upper north facade of WTC7 accelerated on average with a rate approximately equivalent to freefall acceleration for an arbitrarily chosen interval of about 2.25 seconds.

femr2's more accurate and detailed measurements show some point on the roofline was very likely accelerating at significantly > g for one or two brief periods within that interval (early on, actually). This has serious implications for the claim that you want to get at:

• It immediately means that acceleration must have been < g at other periods within that interval, in order for it to work out to = g on average
• In other words, acceleration was variable, not constant, within that interval
• It proves there were forces acting upon the north wall in addition to gravity and structural vertical support
• The preceding point is equivalent to stating that the wall was not in freefall

 

[pgimeno]

One is that when a piece of loose rubble is decelerating it has little to no effect on the other pieces of loose rubble around it.

Wrong. A piece of rubble that is falling at a very close distance above another piece of rubble, will both accumulate the dynamic effect unless the recovery speed is faster than the time the second piece takes to reach the first when it hits. Even in that case, resonance is a factor, and with certain distances the effect can also be that of the sum of the pieces of rubble (minus the loss due to energy dissipation of the spring)

An example of what I am saying would be if you had a beam that would fail with a 100 lb. static load and you dropped five 10 lb. weights on it at different times where each developed a dynamic load of 20 lbs. on impact.

I've highlighted where your problem is. The rubble in the case of the WTC towers was most likely compact enough as for not being considered to fall on the floors at different times, or at the minimum, within very few milliseconds, or piled as described above. That's why in my explanation of dynamic loads I succinctly added that "the volume and density of rubble in the WTC towers probably was more than enough to contribute most of its load as dynamic".

 

[Tony Szamboti]

I'm confused.
Are you suggesting that something, which is in freefall, doesn't speed up?
It does look like you're mixing accelleration and speed like in this post (3450).

What I was trying to say was that its velocity was not dropping during that time and therefore it was was not decelerating. I should have left out the speeding up part as of course it was speeding up while accelerating at freefall.

 

[Tony Szamboti]

Your source correctly states that this static load of 12 floors is equivalent to a load of 6 floors being applied suddenly. Can you explain briefly what they mean by "suddenly"? Is that after dropping from a certain height? What drop distance would that be?

(Hint: I know the answer, just want to see if you know it, too)

Sudden loading is a phenomena which generates up to twice the stress of a static load due the response of the loaded item, if the load is applied suddenly. That is why the NIST FAQ said it would only require 6 floors to fail a floor which could take a static load of 12 floors if the load was applied gradually.

It is not the same as an impact or dynamic load, which can generate many times the static load. 2x the stress is the maximum amplification for sudden loading, and it is generally somewhere between 1 to 2 depending on how quickly the load is applied.

 

[Tony Szamboti]

Wrong. A piece of rubble that is falling at a very close distance above another piece of rubble, will both accumulate the dynamic effect unless the recovery speed is faster than the time the second piece takes to reach the first when it hits. Even in that case, resonance is a factor, and with certain distances the effect can also be that of the sum of the pieces of rubble (minus the loss due to energy dissipation of the spring)



I've highlighted where your problem is. The rubble in the case of the WTC towers was most likely compact enough as for not being considered to fall on the floors at different times, or at the minimum, within very few milliseconds, or piled as described above. That's why in my explanation of dynamic loads I succinctly added that "the volume and density of rubble in the WTC towers probably was more than enough to contribute most of its load as dynamic".

There would be some effect due to proximity and it is on the order of milliseconds that each load would have to be applied with respect to each other to have a cumulative effect. However,the WTC rubble would not have been compacted until a significant number of stories were demolished and that would not have been right away, which is the point of the issue with the missing deceleration and the load required to continue the collapse early on.

Let's see what happens with Dave Thomas' new test, which he seems eager to do, with a bag of rice or sugar and poured rice or sugar. I think the proximity needs to be quite close and that is why I said little to no amplification will occur with the poured material.

 

[NoahFence]

Apparently, the problem for you is I am not wrong and it doesn't matter what some of the obtuse individuals who don't understand that here think.

You think the floors could sustain the extra weight of 12 floors by design. In the unlikely event you're right, did you take into account steel with the consistency of silly putty trying to hold those floors together?

 

[Oystein]

Sudden loading is a phenomena which generates up to twice the stress of a static load due the response of the loaded item, if the load is applied suddenly. That is why the NIST FAQ said it would only require 6 floors to fail a floor which could take a static load of 12 floors if the load was applied gradually.

It is not the same as an impact or dynamic load, which can generate many times the static load. 2x the stress is the maximum amplification for sudden loading, and it is generally somewhere between 1 to 2 depending on how quickly the load is applied.

Ok, good enough, you seem to understand what this "sudden loading" means, although I think it is too vague for non-engineers.

The idea is that of a drop from 0 height. Imagine you are holding the additional free mass an infinitesimally small distance above the intact floor, without actually touching (i.e. without it applying a force on the floor); then you release it "suddenly" (i.e. you stop holding it lifted within 0 seconds, or n infinitesimally short span of time). The "falling" mass will make contact (i.e. start to apply a force) immediately, after zero drop time; since both the floor and the dropped mass are made of real materials, they will compress in responce to this additional force, allowing the center of mass of falling mass to actually drop a finite distance and thus accelerate downwards. The more the materials get compressed, the higher the resisting force, until it equals gravity. From that point on, the falling mass still descends, further compressing materials and increasing force, but now decreasing velocity. If the response is perfectly elastic, then eventually, the mass will stop at a point of maximum compression and upward force. It can be shown that the force now exerted on the lower floor is twice the static load of the falling mass.

In reality, some portion of the response can be assumed to be inelastic (concrete cracking, steel deforming plastically...); this may reduce the peak force from 2*mg to somewhere between 1 and 2 *mg, as you say.
On the other hand, as you say, the falling mass makes contact not with zero velocity, and that greatly increases the max. force.
Further, with collapse initially being asymmetric (yes; tilt and everything), things get complicated rather quickly; locally, forces may be higher or lower than you'd have if the falling mass impacted the floor uniformly throughout.

But now look at another factor: Once the top part of the building descends as a unit, by definition all of its columns are severed, and column ends of top part not in (meaningful) contact with ends of bottom part. Columns ends either bypass floor outside the perimeter, or easily pierce through the floors as they put a great mass on a very small cross section. When the lowest floor still attached to columns in the top part make contact with a floor still attached to the bottom part, both are bolted and welded to their columns with that "12 times static load" strength, so the descending floor doesn't just bring its own mass to the play, but also conveys dynamic forces of the descending columns and all that they are attacjed to, until either the top or the bottom floor-to-column connections break.

Once a floor breaks lose, it has already mostly pancaked with the floor it collided with, and both now descend as a compacted unit onto the next floor below. With now twice the mass, and some additional drop distance, the rext of the tower is a goner.


So I would say: If collapse can't be arrested at the very first floor-to-floor collision, it will never arrest at all. Would you agree?