The 9/11 Forum

T_Szamboti wrote:
Major_Tom wrote:
There is no video of inwardly bowed perimeter columns until the collapse commences.

Does not agree with NIST's report. Indeed NIST did a pattern analysis of the "dimple" to determine its depth.

I didn't write that. It's true for WTC2. WTC1 and the dimple are different.

In both cases early reports of perimeter deformation are big clues. We must start there to track the initiation process.

Major_Tom wrote:

However, if the outer core columns were cut along a face of the building and started dropping they could pull in the perimeter columns through the floors.

But why only along one floor (WTC2)? This is the process that seems to start the collapse of both WTC1 and 2 so early perimeter deformation is very important to understand initiation.

(and it probably ain't sagging floors)


Tony, imagine how the load would redistribute if the 1000 (and 900) row columns ceased to do their job. In both cases it is this row that is along the failed perimeters. The whole side of the building would be totally overstressed. Perimeter failure and leaning is a possibility. When WTC2 leaned, it hinged around an axis corresponding to the 800 or so column row. Clearly the 1000 row gave with (or before??: hint) the perimeter.

Some of these ideas are given in another thread.
(I'll link to it on edit)

Tom, the only thing I will say about the perimeter column failues, due to inward bowing, is that I don't believe for a minute that it was due to sagging floors or that the perimeters all of a sudden gave up themselves. I believe it had something to do with core failures and it is interesting that it happened on the wall in each building that was on the same side as the 1000 row of core columns, but anything beyond what I have said would be unsupported speculation.

David B. Benson wrote:

T_Szamboti wrote:

Major_Tom wrote:There is no video of inwardly bowed perimeter columns until the collapse commences.

Does not agree with NIST's report. Indeed NIST did a pattern analysis of the "dimple" to determine its depth.

Unless you mean that the collapse commenced about 20 minutes before there was any sensible motion of the antenna tower, much less the failure of the exterior walls. But it might be better to call this stage early collapse initiation.

In any case NCSTAR1--6D provides a reconstruction of the state of WTC 1 at around 20 minutes prior to collapse. Of particular interest might be the state of the core columns.

I wrote that there is no video of inwardly bowed perimeter columns until the collapse commences, and I did not stick Tom's name in there like you show it.

It is a fact that no video exists showing inwardly bowed perimeter columns minutes before the collapses occurred. The NIST only has photos which are alleged to show perimeter column bowing for minutes prior to collapse. The fact that NIST does a pattern analysis doesn't say there is video of it occurring.

Major_Tom --- Apologies for my lack of careful eliding. I went back to edit out the mistake.

T_Szamboti --- Please, in the future, to mention which tower you are referring to. I am certainly under the impression that the NYPD helicopter observer actually took video footage of the south wall of WTC 1. If it was actually a still photo, I'd appreciate the correction.

David B. Benson wrote:Major_Tom --- Apologies for my lack of careful eliding. I went back to edit out the mistake.

T_Szamboti --- Please, in the future, to mention which tower you are referring to. I am certainly under the impression that the NYPD helicopter observer actually took video footage of the south wall of WTC 1. If it was actually a still photo, I'd appreciate the correction.

I am actually talking about both towers here. You are welcome to try and find video of the perimeter columns bowing inward for minutes prior to collapse. I don't think you will find any. I have heard anecdotally that someone in a helicopter said they thought the North Tower was going to collapse because the columns were bowing inward. That isn't video.

David, it looks like the dust zone stays in the middle of the collapse. if h(t) is the roofline as function of time then that zone is at h(t)/2, but dust is hard to measure from video.

T_Sz, dr. G, I also though that there is deceleration before the top section hits the ground, but this is something that should be looked at in more detail, maybe if OWE is back he can have a look

Einsteen and T. Sz:

I think we must be very careful about using the word decelerate when we really mean less acceleration! Indeed, many physics textbooks advise never to use the word "decelerate" but recommend instead using "negative acceleration" to avoid the ambiguity inherent in the word "decelerate".

In the example of the French building I posted yesterday, which no one has tried to calculate, (too bad, because it's very instructive), you have a good example of decreasing acceleration accompanied by a velocity increase. Thus if the upper block descends the first 7 meters in 1.22 seconds and the first 14 meters in 1.78 seconds, the drop from 7 to 14 meters obviously took 0.56 seconds. This implies that the acceleration changed from 9.4 m/s^2 to 3.69 m/s^2 at the 1.22 second mark. However it's simple to show that the velocity would have been 11.5 m/s at 1.22 seconds and 13.53 m/s at 1.78 seconds, so we should not say the upper block decelerated, because the velocity was always increasing, but rather there was a reduction in the acceleration.

And, by the way, this example nicely illustrates how precise the video measurements must be to detect these effects! After 1.22 seconds, or 7 meters of "near free fall" at 9.4 m/s^2, the block is moving at 11.5 m/s. It turns out that there is a change by a factor of two in the acceleration if the time to drop the 2nd 7 meters changes by a mere 0.04 seconds! It follows that in order to detect a jolt at 1.22 seconds (in my example!), the drop data must be recorded with a precision of better than 0.1 meters which implies a sampling rate of 10 milliseconds. I seriously doubt this level of precision is possible with any of the videos we are discussing.

A few additional points on the demolition of the ABC (Balzac) building in France:

By far the best video I have seen of this demolition is at:

http://www.strimoo.com/video/12509995/M ... ideos.html

See especially the excellent "slo-mo" clip starting at 2:31 into the video.

Apparently the French call this (non-explosive) demolition technique verinage.

In this particular case the hydraulic actuators were on the 7th floor

Thanks for the lessons dr.G, I looked in my dictionary and that says "vaart (doen) verminderen" which means that the speed is decreasing, so you are right that we should not use that word. but negative acceleration is also not right if it is still positive.

Great video!

I created a curve of the left front corner


http://i40.tinypic.com/oa55iv.jpg

It looks very linear and the part at the end is due to the tilting, the roof's colour on the video has a lighter gray shade, but the dark gray shade's curve is very linear.

Einsteen:

I knew you would like that video - trust the French to set it to some nice music .....

And thanks for the smearogram!

Pardon the pun, ..... but can you blow it up!

Or do a drop vs. time plot.

Very roughly I estimate a drop of about 6.5 meters at 1.15 seconds and 15 meters at 2.0 seconds.

However, right now, I SEE NO WHAM!

Dr. G wrote:Einsteen and T. Sz:

I think we must be very careful about using the word decelerate when we really mean less acceleration! Indeed, many physics textbooks advise never to use the word "decelerate" but recommend instead using "negative acceleration" to avoid the ambiguity inherent in the word "decelerate".

In the example of the French building I posted yesterday, which no one has tried to calculate, (too bad, because it's very instructive), you have a good example of decreasing acceleration accompanied by a velocity increase. Thus if the upper block descends the first 7 meters in 1.22 seconds and the first 14 meters in 1.78 seconds, the drop from 7 to 14 meters obviously took 0.56 seconds. This implies that the acceleration changed from 9.4 m/s^2 to 3.69 m/s^2 at the 1.22 second mark. However it's simple to show that the velocity would have been 11.5 m/s at 1.22 seconds and 13.53 m/s at 1.78 seconds, so we should not say the upper block decelerated, because the velocity was always increasing, but rather there was a reduction in the acceleration.

And, by the way, this example nicely illustrates how precise the video measurements must be to detect these effects! After 1.22 seconds, or 7 meters of "near free fall" at 9.4 m/s^2, the block is moving at 11.5 m/s. It turns out that there is a change by a factor of two in the acceleration if the time to drop the 2nd 7 meters changes by a mere 0.04 seconds! It follows that in order to detect a jolt at 1.22 seconds (in my example!), the drop data must be recorded with a precision of better than 0.1 meters which implies a sampling rate of 10 milliseconds. I seriously doubt this level of precision is possible with any of the videos we are discussing.

It isn't just reduced acceleration in the abc tower, although that does occur after the first two story fall probably due to the loose debris of the demolished floors, there was a real 15 to 20% velocity loss after the third story of the fall. This would have been when the upper block impacted the intact lower block. In the terms you would like to use, that means there was negative acceleration. It is quite unambiguous also, as it takes time for the velocity to recover and the velocity drop can be detected with more than one data point. Measure it and see for yourself.

In WTC 1 the energy requirements to buckle the columns on the 97th and 99th floors would have drained 85% of the pre-impact velocity and conservation of momentum in picking up the 98th floor slab would have reduced it further. It would have been a dramatic velocity loss if there was an impulse after a 12 foot drop. There would then have been about a 1000 millisecond window to detect the velocity loss while it was recovering to pre-impact levels. But we don't see this velocity loss in WTC 1, therefore there was no negative acceleration, which means no force amplification and no natural mechanism for collapse. There is no way around that and that is why Dr. Bazant thought there had to be an impact and velocity loss or negative acceleration.

T. Szamboti:

I find it interesting that you apparently want to compare the collapse of WTC 1 with the collapse of the ABC Tower in France, and consider one to be "natural" and the other, presumably, to be "unnatural". The question is: which is which?

But be that as it may, when you say there was negative acceleration in the ABC Tower collapse, I am not so sure. And I have measured the collapse rate over the first 2.5 seconds and all I can say is that a jolt is possibly present, but only as a reduction in downward acceleration after 7 meters of drop. However, negative acceleration is never present as far as I can tell.

What I find most interesting right now is how the ABC Tower was "pulled" by those clever French engineers, and the possibility that there may have been an acceleration greater than g over the first 7 meters. Now this is reminiscent of WTC 7

But this shows that the ABC demolition was far from being "natural", thus I would not expect WTC 1 to replicate the ABC collapse - on the contrary, I would expect it to be quite different!

What I find most interesting right now is how the ABC Tower was "pulled" by those clever French engineers, and the possibility that there may have been an acceleration greater than g over the first 7 meters. Now this is reminiscent of WTC 7

Exactly.

The abc demo is a useful experiment for us.

A more detailed smearogram of the abc collision would teach us a lot. Wonderful comparison with WTC7.

Looking for velocity loss (or just plotting downward speed to see what happened) near the 7 meter mark seems to tell us all we need to know.

We know a collision is taking place. A big one. Interesting to see if it is detectable in the smearogram.

It isn't just reduced acceleration in the abc tower, although that does occur after the first two story fall probably due to the loose debris of the demolished floors, there was a real 15 to 20% velocity loss after the third story of the fall.

But be that as it may, when you say there was negative acceleration in the ABC Tower collapse, I am not so sure. And I have measured the collapse rate over the first 2.5 seconds and all I can say is that a jolt is possibly present, but only as a reduction in downward acceleration after 7 meters of drop. However, negative acceleration is never present as far as I can tell.

We have to find out how a real head-on collision appears as data. This is our opportunity. It will help us understand the WTC7 fall.

(WTC7 vs ABC data and smearograms would be a great topic for a paper. A real eye-opener.)

Dr.G, It will be a job to translate it in real h(t) because of the perspective and error margins etc. maybe a pix(t) function is a good starter.. I’ll later create a data set if you like.

All,
I think that this collapse is interesting, not because of the simultaneous symmetrical initiation but because of the fact that this one of the few videos that potentially could be used to check the crush-down crush-up theory. wtc7 also didn’t start completely at the bottom but there are only videos of the top part. I recently did some primitive experiments at home with big plastic lego blocks which I didn’t really close (by stacking them but not pushing them together) and if you drop that on the floor the blocks at the bottom first are pushed together as expected, in this equal mass/force situation. But a crush-down is a totally different experiment, I couldn’t really reproduce it, although scaling is relevant it seems very sensitive to the difference in friction forces, but as expected if you place a stack of them and push the upper block down with your hand they all are pushed together at the same time and I thought that also happened by dropping a heavy object on it. In a huge mass situation the inertia of the blocks is relevant and will affect the effect but I was amazed that the crush-up worked very well with those light blocks.

The video, of course, should be studied in more detail, but my first temporary impression is that
- the tilt of the top section (it started in fact non-tilted) becomes relevant after some collapse time and is due to the relevant resistance of the material
- the top section is absolutely not intact during the “crush down”
- there are no violent ejections like wtc1,2 only some dust.