Merged Discussion of femr's video data analysis

[femr2]

You have a smoothing method better than Savitzky-Golay ?

Yes

Then what is it ?

 

[femr2]

Do you have an issue with the results of the 16 storey ANSYS analysis and how would you apply them to the LS-DYNA model?

An issue ? No, many issues. Probably not the right thread for that discussion.

For reference, the development of the kink (t=0) is at 16 seconds on that graph.

No. If you mean the NIST North facade T=0, then the *kink* is a result of perspective distortion from the Cam#3 viewpoint, and is primarily N-S flexure of the facade.

The NIST Global T=0 is defined by the formation of the kink in the East penthouse roofline, which occurs around the 12s mark in the global simulation.

 

[Reactor drone]

An issue ? No, many issues. Probably not the right thread for that discussion.


No. If you mean the NIST North facade T=0, then the *kink* is a result of perspective distortion from the Cam#3 viewpoint, and is primarily N-S flexure of the facade.

The NIST Global T=0 is defined by the formation of the kink in the East penthouse roofline, which occurs around the 12s mark in the global simulation.

Seems surprising that NIST would say 16 s then,

"Two timing reference frames are used when discussing global model results: A calculation reference frame and a collapse reference frame. the LS-DYNA calculation reference frame is described in Section 12.3.2. the collapse reference frame is absed on observation times from photographic and videographic records (Chapter 5). The collapse reference time of 0.0 s starts when a kink was observed in the roofline of the east penthouse, as viewed from the north. In the global analysis results, the kink in the roofline of the east penthouse occurred at 16.0 s in the calculation reference time."

 

[Myriad]

Connection failures might be the cause of buckling. But connection failures without buckling do not cause movement.

Wind causes movement So does heat. So does...

...earthquakes, so does a collision with a derailed speeding freight train, so does invisible dwarves with hydraulic jacks...

We're still talking about building 7, right? Do you think the movement was caused by wind? If so, then please state the hypothesis. If not, why bring it up?

I've already addressed thermal expansion as a direct cause of the movement, and the obvious reason why I think it is unlikely. If you have a scenario in mind by which heating or cooling of members could have occurred at the necessary rate to cause the movements observed, please describe it. Otherwise, I will continue to discount the possibility. (Though it remains a good candidate for earlier, slower movements also observed, e.g. the formation of the "bulge" hours earlier.)

So you consider all flexing motion to be the result of buckling. Interesting.

All the demonstrated flexing motion that's currently under discussion that occurred in building 7 just prior to collapse, yes.

Because only buckling releases energy in a way that can do work on the rest of the structure, and work is necessary to cause movement or distortion.

I assume this is linked to your statement above. Hmm. Rather boxing yourself in don't you think ?

Yes, it's bold and presumptuous of me to have proposed an actual falsifiable hypotheses and stated it clearly. I'm putting myself in grave danger of being shown wrong about a miniscule historical detail. But that's how I roll.

Something in the structure descended vertically, to provide the energy to cause the oscillation.

Why not lateral restraint ?

"Lateral" and "restraint" are two very useful words. But, without specifying lateral restraint of what, by what, causing what, they do not amount to an alternative hypothesis of what did the work upon the structure to make it move.

That doesn't necessarily mean the specific small part you're measuring descended vertically (except in the sense that unless members are able to stretch or have become detached from the ground, horizontal movement without some small amount of vertical movement is geometrically impossible -- but that amount would probably be too small to directly observe).

Do you think it would be possible to detect the effect of wind upon a 47 storey structure ?

Yes. Again, is your hypothesis that the early movement you measured was caused by wind? If so, let's hear it. If not, then why are you wasting your own time bringing it up?

I'll ask again, what event or events physically define a release point?

As you already know T=0 is subjective, but is most simply put as the point in time at which, in this context, unrecoverable vertical motion begins. Pretty simple really.

Okay, moving on.

How can you argue about the timing of an event you have not defined?

It's been defined many times

T=0 is subjective

Within the available margins of error, yes.

What margin ? For the North facade, or that central location, or the NW corner, or the building ?

Margins of error apply to measurement methods, not to building locations. Here's one obvious and significant source of a margin of error:

T=0 is subjective

Because I realize I've been wasting time discussing the timing of an event that is not physically defined.

Release point has been defined many times. Want a list ?

T=0 is subjective

Start with a circular definition (the release point is T=0, T=0 is set at the release point) and it's no surprise to get a circular argument. Define the release point physically, and we can continue that discussion.

See above, again.

T=0 is subjective

You have to bear in mind you are choosing the NIST data, which suffers from numerous perspective related distortions and issues...

T=0 is subjective

Because the NIST T=0 defines the beginning of Stage 1.

It's meaningless. That point was in motion long beforehand. A point I HOPE you don't ignore.

T=0 is subjective

NIST T=0 is slightly too early.

Let's put this one on a repeat loop.

NIST T=0 is slightly too early.

T=0 is subjective

NIST T=0 is slightly too early.

T=0 is subjective

NIST T=0 is slightly too early.

T=0 is subjective

I think that concludes the discussion on NIST's (and your) placements of T=0. What else need be said?

The timing of Stage 2 is defined by different physical events (from the observed velocity-vs-time curve becoming approximately linear, until it is no longer so) and would not be affected at all by moving the time marker of the beginning of Stage 1.

We can revert to near-at-over g discussion shortly I am sure.

I'll wait.

Respectfully,
Myriad

 

[femr2]

Seems surprising that NIST would say 16 s then,

Aha. They are referring to the other global model. Both values are correct.

What is your point ?

Both models begin collapsing immediately after application of *ANSYS damage*.

 

[femr2]

Do you think the movement was caused by wind?

No.

If not, why bring it up?

To highlight that there are many causes of motion which are not the result of buckling.

All the demonstrated flexing motion that's currently under discussion that occurred in building 7 just prior to collapse, yes.

I note you are restating *just prior to* again. 6 minutes prior to ?

Yes, it's bold and presumptuous of me to have proposed an actual falsifiable hypotheses and stated it clearly. I'm putting myself in grave danger of being shown wrong about a miniscule historical detail. But that's how I roll.

This is not about immature point scoring. It is about correctly and accurately determining the cause of the very early motion, then comparing that with the proposed NIST assertions.

Have you read appendix C yet ?

"Lateral" and "restraint" are two very useful words. But, without specifying lateral restraint of what, by what, causing what, they do not amount to an alternative hypothesis of what did the work upon the structure to make it move.

I've suggested fracture of structural components (rather than buckling) several times now, which reduce lateral restraint upon the building low down, and allow much greater than normal flexure, especially given the large gash on the south side around column 21.

Margins of error apply to measurement methods, not to building locations.

No stated margin of error then ?

I'd be comfortable suggesting a +/-0.1s margin to the NW corner T=0.

I think that concludes the discussion on NIST's (and your) placements of T=0. What else need be said?

NISTs North facade T=0 is out by about a second, as repeatedly highlighted.

That makes a significant difference to their stated 40% above freefall assertion, along with the descent time.

 

[Oystein]

Do you think the movement was caused by wind?

No.

So are you saying that wind is not among the possible causes of motion here? Okay.

If not, why bring it up?

To highlight that there are many causes of motion which are not the result of buckling.

But you just rules wind out! Question remains: Why bring up wind, if not even you believe that wind caused that lateral movement?

You see there is a sentence you cut from Myriad's text between the two sentences you quoted above. May I remind you of what it was?

If so, then please state the hypothesis.

I suggest you do just that: State your hypothesis of what caused the lateral movements you observed earlier. Where did the energy come from, if nor vertical movement and wind?

It is about correctly and accurately determining the cause of the very early motion, then comparing that with the proposed NIST assertions.

NIST has a hypothesis. If you want to compare it to anything, that better be a hypothesis, too.
So what is your hypothesis?

I've suggested fracture of structural components (rather than buckling) several times now, which reduce lateral restraint upon the building low down, and allow much greater than normal flexure, especially given the large gash on the south side around column 21.

But Myriad asked you to be specific on that. May I remind you of what he suggested:

"Lateral" and "restraint" are two very useful words. But, without specifying lateral restraint of what, by what, causing what, they do not amount to an alternative hypothesis of what did the work upon the structure to make it move.

No stated margin of error then ?

I'd be comfortable suggesting a +/-0.1s margin to the NW corner T=0.

Is that a guess, or does this follow from any math?

NISTs North facade T=0 is out by about a second, as repeatedly highlighted.

T=0 is subjective, as repeatedly highlighted.

 

[Reactor drone]

Aha. They are referring to the other global model. Both values are correct.

What is your point ?

Both models begin collapsing immediately after application of *ANSYS damage*.

Still don't know where you're getting 12 s from,

The first model (section 12.4.4), the one you provided the graph for, with 4 hours fire damage added collapsed at 16 s calculation time

The second model (section 12.4.5), with 3.5 hours fire damage added didn't collapse at all.

The third model (section 12.4.6), without debris damage starts to collapse at 13.3 s calculation time.

The forth model (section 12.4.7), with no damage apart from removal of part of column 79 starts to collapse at 5.3 s calculation time.

Since the LS-DYNA model can't do fire damage over time but ANSYS can the whole point was to bring the building condition to a known timepoint in LS-DYNA and let it run from there. As you can see application of ANSYS damage didn't always cause collapse making your comment "Well, it's useful to understand that their descent simulation was based upon a model which was never going to do anything except collapse..." a little misleading.

 

[femr2]

Still don't know where you're getting 12 s from

From the actual NIST video model animation.

Since the LS-DYNA model can't do fire damage over time but ANSYS can the whole point was to bring the building condition to a known timepoint in LS-DYNA and let it run from there. As you can see application of ANSYS damage didn't always cause collapse making your comment "Well, it's useful to understand that their descent simulation was based upon a model which was never going to do anything except collapse..." a little misleading.

Not at all. The other models you mention are not the one I'm talking about, as you clearly are aware. The only one you listed that didn't immediately collapse had no damage applied.

My point being clear...the model began collapsing immediately after it was allowed to move.

Interesting to note the role of the damage in the model.

How do you think they got to a point where damage had been identified, but the thing hadn't collapsed ?

Are you fully clear on the relationship between ANSYS and LS-DYNA ?

 

[femr2]

So are you saying that wind is not among the possible causes of motion here? Okay.

No. Wind will be a part of it, but not the primary factor, of course.

But you just rules wind out!

Incorrect. I simply answered that I don't think wind was the cause of (all of the) movement.

Ruled wind out just by saying so ? Hmm.

You see there is a sentence you cut from Myriad's text between the two sentences you quoted above. May I remind you of what it was?

Do as you please

May I remind you of what he suggested

Again, you can do whatever you please. I can read by the way No need for you to reiterate things I've already read.

Is that a guess, or does this follow from any math?

It's an assertion based upon close inspection of the displacement/velocity/acceleration profiles.

T=0 is subjective, as repeatedly highlighted.

Indeed. I stated such

The NIST North facade T=0 is out by about 1s.

 

[Oystein]

T=0 is subjective, as repeatedly highlighted.

Indeed. I stated such

The NIST North facade T=0 is out by about 1s.

<- *bangs head*

 

[twinstead]

<- *bangs head*

Indeed.

Ergo you are only interested in a perpetual debate, not on finding any common ground or "truth".

 

[Reactor drone]

Aha. They are referring to the other global model. Both values are correct.

What is your point ?

Both models begin collapsing immediately after application of *ANSYS damage*.

From the actual NIST video model animation.


Not at all. The other models you mention are not the one I'm talking about, as you clearly are aware. The only one you listed that didn't immediately collapse had no damage applied.

My point being clear...the model began collapsing immediately after it was allowed to move.

Interesting to note the role of the damage in the model.

How do you think they got to a point where damage had been identified, but the thing hadn't collapsed ?

Are you fully clear on the relationship between ANSYS and LS-DYNA ?

You posted a graph showing the set up of the first LS-dyna model with 4 hours damage applied stating that it collapsed immediately after the application of the ANSYS damage. This graph only matches the first two cases since they're the only ones where impact damage was applied. I posted a section from the NIST report showing that the start of global collapse occured at 16s on the graph and you changed your story from immediate collapse to collapse at the 12s mark then claimed I was talking about "another" model and that both 12 and 16 seconds were correct. I then posted up the times of all the models, none of which have t=0 being at 12 s calculation time so now you're claiming you're pulling the time off the animation.

Just out of interest which animation are you using, the one with fire and impact damage that matches the graph you posted or the one with only fire damage that doesn't?

Do you expect a model of a building with no damage to collapse? Since NIST was studying the collapse do you think having a model with realistic fire damage applied (from the ANSYS model) might help shed light on the collapse mechanism?

The 3.5 hour damage model that didn't collapse did have ANSYS damage applied but the damage wasn't enough to bring about a collapse.

The models are build to study the collapse but the method of loading the gravity, impact damage and fire induced damage has no bearing on whether or not the simulation collapses, only the extent of the damage has an effect.

 

[Oz1976]

...earthquakes, so does a collision with a derailed speeding freight train, so does invisible dwarves with hydraulic jacks...

We're still talking about building 7, right? Do you think the movement was caused by wind? If so, then please state the hypothesis. If not, why bring it up?
I've already addressed thermal expansion as a direct cause of the movement, and the obvious reason why I think it is unlikely. If you have a scenario in mind by which heating or cooling of members could have occurred at the necessary rate to cause the movements observed, please describe it. Otherwise, I will continue to discount the possibility. (Though it remains a good candidate for earlier, slower movements also observed, e.g. the formation of the "bulge" hours earlier.)

We already know his answer to the hilited Myriad. The first part requires a certain explosive known as a hush-a-boom. The second answer as to why he wont describe it is he's trying to appear scientific in this nonsense. If one were to take a look at the videos he has posted on his youtube channel his beliefs are clear. All the collapse videos are listed as "World Trade Center Demolition".

 

[femr2]

You posted a graph showing the set up of the first LS-dyna model with 4 hours damage applied stating that it collapsed immediately after the application of the ANSYS damage.

Which it did.

This graph only matches the first two cases since they're the only ones where impact damage was applied.

Therefore excluding such.

I posted a section from the NIST report showing that the start of global collapse occured at 16s on the graph and you changed your story from immediate collapse to collapse at the 12s mark then claimed I was talking about "another" model and that both 12 and 16 seconds were correct.

I included the other available global collapse animation, absolutely. Changed my story ? Nope. Accepted both values, absolutely. Both models collapsed.

I then posted up the times of all the models, none of which have t=0 being at 12 s calculation time so now you're claiming you're pulling the time off the animation.

I was never claiming anything but. Just because you didn't know the source of the value doesn't mean the source has changed, y'know

Just out of interest which animation are you using, the one with fire and impact damage that matches the graph you posted or the one with only fire damage that doesn't?

Obviously, we're discussing T=0 of both at the mo.

http://www.youtube.com/watch?v=7-nbfeGjwZU
http://www.youtube.com/watch?v=c600UompC-I
http://www.youtube.com/watch?v=elnfqCMjFS8
http://www.youtube.com/watch?v=TNEKtvB80us

Do you expect a model of a building with no damage to collapse?

Irrelevant question. Do you ?

The models are build to study the collapse but the method of loading the gravity, impact damage and fire induced damage has no bearing on whether or not the simulation collapses, only the extent of the damage has an effect.

The method ? What do you mean ? As soon as the thing was *allowed* to move, it collapsed there and then.

My prior questions...

How do you think they got to a point where damage had been identified, but the thing hadn't collapsed ?

Are you fully clear on the relationship between ANSYS and LS-DYNA ?

How long after application of *ANSYS damage* did collapse ensue ? (no veering off into other models, thanks)

Did the model run for a few hours of simulated time then collapse, or did it immediately crumple like a overgrown tin can ? (no veering off into other models, thanks)

Did collapse ensure before application of *ANSYS damage* ? (no veering off into other models, thanks)

 

[beachnut]

Which it did.

Therefore excluding such.

I included the other available global collapse animation, absolutely. Changed my story ? Nope. Accepted both values, absolutely. Both models collapsed.

I was never claiming anything but. Just because you didn't know the source of the value doesn't mean the source has changed, y'know

Obviously, we're discussing T=0 of both at the mo.

http://www.youtube.com/watch?v=7-nbfeGjwZU
http://www.youtube.com/watch?v=c600UompC-I
http://www.youtube.com/watch?v=elnfqCMjFS8
http://www.youtube.com/watch?v=TNEKtvB80us

Irrelevant question. Do you ?

The method ? What do you mean ? As soon as the thing was *allowed* to move, it collapsed there and then.

My prior questions...

How do you think they got to a point where damage had been identified, but the thing hadn't collapsed ?

Are you fully clear on the relationship between ANSYS and LS-DYNA ?

How long after application of *ANSYS damage* did collapse ensue ? (no veering off into other models, thanks)

Did the model run for a few hours of simulated time then collapse, or did it immediately crumple like a overgrown tin can ? (no veering off into other models, thanks)

Did collapse ensure before application of *ANSYS damage* ? (no veering off into other models, thanks)

I had it on good word you were going to wait to spring this revelation on us in September.

Thinking about your work related to 9/11 Conspiracy Theories, located here in the Conspiracy Theories section. Do you have a goal and conclusion for this terrific ground breaking work? How does it fit with your "demolition" theory on the WTC? With respect to your 62 videos on You-Tube, titled "Demolition", will this super video analysis help or hurt your theory the WTC complex suffered demolition at the hands of unknown people? Where do is your effort with respect to your claim, "fictional official theory"?

I found after seeing your work and all the available evidence, WTC 7 fell due to fire. What do you think made WTC 7 collapse, the big picture? The big question I have, does your work support your demolition theory?

 

[Major_Tom]

I had it on good word you were going to wait to spring this revelation on us in September.

Thinking about your work related to 9/11 Conspiracy Theories, located here in the Conspiracy Theories section. Do you have a goal and conclusion for this terrific ground breaking work? How does it fit with your "demolition" theory on the WTC? With respect to your 62 videos on You-Tube, titled "Demolition", will this super video analysis help or hurt your theory the WTC complex suffered demolition at the hands of unknown people? Where do is your effort with respect to your claim, "fictional official theory"?

I found after seeing your work and all the available evidence, WTC 7 fell due to fire. What do you think made WTC 7 collapse, the big picture? The big question I have, does your work support your demolition theory?

Thank you for making a post without using the word "dovetail".

 

[Reactor drone]

Which it did.

Provided you count the beginning of internal failures as a building collapsing or count +7.5s as being immediately.

Therefore excluding such.

I included the other available global collapse animation, absolutely. Changed my story ? Nope. Accepted both values, absolutely. Both models collapsed.

For accuracy the values for the models you're using are 16 seconds and 13.2 seconds calculation time for the start of global collapse.

I was never claiming anything but. Just because you didn't know the source of the value doesn't mean the source has changed, y'know

Well I figured since you're using the NIST animations you'd be using their figures rather than some approximation you worked out from a grainy youtube video.

Obviously, we're discussing T=0 of both at the mo.

http://www.youtube.com/watch?v=7-nbfeGjwZU
http://www.youtube.com/watch?v=c600UompC-I
http://www.youtube.com/watch?v=elnfqCMjFS8
http://www.youtube.com/watch?v=TNEKtvB80us

The graph you posted and the accompanying T=0 figure applies only to the first model shown.

Irrelevant question. Do you ?

The method ? What do you mean ? As soon as the thing was *allowed* to move, it collapsed there and then.

It was allowed to move when gravity was applied and it was allowed to move when the damage was applied. It was also allowed to move after the ANSYS damage was applied and in one simulation it failed to collapse therefore application of damage and being allowed to move are not always the beginnings of collapse, only when damage is severe enough does the internal collapse start so whatever problem you have with how damage was applied seems to be based on a misunderstanding of the models.


My answers in bold below.

My prior questions...

How do you think they got to a point where damage had been identified, but the thing hadn't collapsed ?From the 16 storey ANSYS analysis

Are you fully clear on the relationship between ANSYS and LS-DYNA ? Yes

How long after application of *ANSYS damage* did collapse ensue ? (no veering off into other models, thanks) Internal failures started after application of damage with the start of global collapse(T=0) occuring 7.5 seconds later

Did the model run for a few hours of simulated time then collapse, or did it immediately crumple like a overgrown tin can ? (no veering off into other models, thanks) Hours of fire damage had already been simulated in ANSYS so at the beginning of the simulation the LS-DYNA model was already at the 4 hour fire damage point.

Did collapse ensure before application of *ANSYS damage* ? (no veering off into other models, thanks) Obviously not, hence my earlier question about whether you thought an undamaged model would collapse.

 

[femr2]

Provided you ... count +7.5s as being immediately.

Of course. The model was in a state that immediately colapsed. Do you dispute this point ?

For accuracy the values for the models you're using are 16 seconds and 13.2 seconds calculation time for the start of global collapse.

Where is your 13.2s figure determined from ? *What NIST said ?* Inspection of their released video animations shows motion of the East Penthouse roofline before that point. Not a big deal, but always worth actually cross-checking between published numerical values and the actual animations they are derived from.

Well I figured since you're using the NIST animations you'd be using their figures rather than some approximation you worked out from a grainy youtube video.

The *grainy YT video* is the released NIST animation in all it's glory. Did NIST release particularly poor quality instances of their video animations ? Absolutely. The *blame YT* thing doesn't wash. If you are aware of instances of the animation in higher quality, by all means direct me to them.

The graph you posted and the accompanying T=0 figure applies only to the first model shown.

Nope. First and last video contain the specified animation.

whatever problem you have with how damage was applied seems to be based on a misunderstanding of the models.

I fully understand the reasoning behind the setup procedure. I do have issue with the way damage was applied of course. Damage from one method determined via *a procedure* NIST made up by which they removed members by some arbitary criteria from the model input into LS-DYNA. The starting state of the model in LS-DYNA collapsed immediately.

How do you think they got to a point where damage had been identified, but the thing hadn't collapsed ?From the 16 storey ANSYS analysis

Yes, but why hadn't it already collapsed in the ANSYS Wokbench ?

Are you fully clear on the relationship between ANSYS and LS-DYNA ? Yes

Hmm...
http://www.internationalskeptics.com/forums/showthread.php?p=4675487&highlight=ansys#post4675487

Did the model run for a few hours of simulated time then collapse, or did it immediately crumple like a overgrown tin can ? (no veering off into other models, thanks) Hours of fire damage had already been simulated in ANSYS so at the beginning of the simulation the LS-DYNA model was already at the 4 hour fire damage point.

I'd call that dodging the question.

Did collapse ensure before application of *ANSYS damage* ? (no veering off into other models, thanks) Obviously not

So would you agree that application of (aka removal of significant sections of the building) *ANSYS DAMAGE* altered the state of the model from one which would have remained standing to one that immediately collapsed ?

 

[NoahFence]

Where'd MT go? I sure do miss his well thought out, insightful posts.