Experts Can't Know Everything

[gumboot]

I've noticed an interesting feature of many conspiracy theorist arguments lately.

Anyone who has expertise in a given area is also very familiar with the limits of that expertise. By this I am not referring to what they do not know, but what they know they cannot know.

There are limitations on science. Given a specific set of evidence, only so much can be learned. This is a vital piece of knowledge for an expert to be aware of.

Conspiracy Theorists simply do not understand this.

We see it often in regards to NIST's study of the WTC collapses. They have stated repeatedly that it is simply not possible to model the actual collapse progression, as the number of variables is too high.

Conspiracy Theorists cannot get their heads around this.

Likewise, recently I have been discussing with a poster who attempted to analyse the Pentagon security footage to determine the size of the captured object - concluding it had a diameter less than 2ft.

When I pointed out the flaws in his analysis, he demanded to know how I would determine the size. He seemed totally incapable of getting his head around the idea that you cannot calculate the size.

Likewise with the footage of UA175 hitting WTC2, I have pointed out that the margin of error is so large that calculating the deceleration of the aircraft is impossible.

It is not about lacking expertise. It is a simple matter of being absolutely impossible. It cannot be done. Insufficient data exists. It would be like taking a black and white photo of a house, and then demanding that a photographic expert tells you what colour the door is.

Does anyone else have any experience with Conspiracy Theorists failing to grasp the limits of what can be determined?

-Gumboot

[TriskettheKid]

I'd be careful throwing around the word "impossible."

But otherwise, agreed.

[rwguinn]

I'd be careful throwing around the word "impossible."

But otherwise, agreed.

gummy was very careful.
Likewise with the footage of UA175 hitting WTC2, I have pointed out that the margin of error is so large that calculating the deceleration of the aircraft is impossible.
the key there is the bolded part.:
Given: Trisketthekid is at least 5 feet tall. Gumboot is at least 4 feet tall. Who is taller: Gumboot or trisketthekid?
Whit the information at hand, the taks is impossible. Any result is due to pure guesswork. The margin of error is way too big. Even some PhD's fail to realize this fact.

[gumboot]

rwguinn has it spot on, this is about knowing the limitations of the evidence you currently possess.

Additional evidence will of course allow for additional information to be determined. For example if sensors were located at spaces of every foot all over the WTC, information about the collapse progression could be mapped. Likewise, if a high definition high frame rate camera were pointed at the WTC at the same level as UA175, and with the camera positioned 90 degrees to UA175's flight trajectory (a set up much like the footage of the F4 hitting the concrete wall) there would be sufficient information to make a much more accurate calculation.

-Gumboot

[Bell]

Lucky the troofers can rely on common sense instead of science and accurate calculations.

[DavidJames]

For many, the more you know, the more you realize you don't know. For CTists, the less they know, the more they think they know.

[Bell]

Is it arrogance or stupidity to think you know it better than the experts? For example, non of the egineers in the world (except the oddball or two) thinks the WTC was a CD. But nooo, a kid and his computer know better.

[BenBurch]

I lay the blame squarely on the school systems. Science Education barely exists in many schools.

[Mel Odious]

Oh, the engineers of the world do all know the buildings were blown up. They were just all paid off by the Bush administration, the N.W.O., and other shadowy clandestine organizations to keep them from talking about it.

The next time you attend a college graduation, you might even see all the new engineering grads being pulled aside and handed envelopes stuffed with hush money. It's a pain in the neck sending agents to hundreds of colleges around the world every single year, not to mention the obvious difficulties of coming up with huge piles of cash serruptitiously, but hey, secrets are secrets and must be kept.

[Sabrina]

I personally find it irritating as all get out when a twoofer presumes to tell me something about a situation I've been involved with and they know absolutely nothing about. But as that's not exactly relevant to the OP, I'll move on.

I've noticed much the same thing with the TM in the (relatively) short amount of time I've spent researching 9/11; it's like they can't wrap their minds around the idea that the universe is ultimately unknowable. They think we should be able to accurately determine the exact detail of everything having to do with 9/11, from the collapse of the towers all the way down to the exact reason why Atta had what he had in his luggage, and it's just not possible, at least with our current level of technology. Now, if we had like a Star Trek:The Next Generation level of technology, we'd have a slim chance of doing so, but even then I'd be doubtful. It might, MIGHT be because a large margin of twoofers are young teenagers. At that age, you think you know everything and everyone else around you is stupid; it's not until you gain maturity that you begin to think critically about things. That would be my assessment from my thankfully dimming memories of high school and a degree in psychology, by the way. But it's not all the age thing either; some people are just more prone to thinking more with their emotions instead of their logical capabilities, and then THEY'RE the ones who can't understand why WE don't agree with THEM. From my admittedly layman's assessment of the TM, it is comprised largely of those prone to mild psychological disorders such as a tendency toward paranoia and those who are simply not mature enough to make a critical, logical assessment of the world and realize that there are many many MANY things that are currently out of our realm of knowledge. The first group pretty much has no hope; they'll be that way for good unless they seek help, but at least the second group has a chance to change, and ultimately that, I think, is what those of us on the debunking side are working for.

[Belz...]

Likewise, recently I have been discussing with a poster who attempted to analyse the Pentagon security footage to determine the size of the captured object - concluding it had a diameter less than 2ft.

That would be Terral, right ?

[Belz...]

At that age, you think you know everything and everyone else around you is stupid; it's not until you gain maturity that you begin to think critically about things.

Well, that rings a bell !

[geni]

It would be like taking a black and white photo of a house, and then demanding that a photographic expert tells you what colour the door is.


For a high enough quality photo there are a couple of ways that might be posible.

[CHF]

Twoofers play the same game that creationists do: If something cannot be explained or modeled then that must mean God did it.

Likewise, the ability to somehow model the dynamics of the 110 story building collapse means that it must have been a demolition.

It's how people argue when they have no evidence.

[CHF]

They think we should be able to accurately determine the exact detail of everything having to do with 9/11, from the collapse of the towers all the way down to the exact reason why Atta had what he had in his luggage, and it's just not possible, at least with our current level of technology. Now, if we had like a Star Trek:The Next Generation level of technology...

And there's the problem: twoofers think we do have that kind of technology.

[Sabrina]

And that more than ever indicates to me the lack of connection to reality that twoofers demonstrate.

[afinemadness]

I have read this study before and it explains some of the thought patterns of twoofers. It is not a direct correlation but you will see them in it.
http://www.apa.org/journals/features/psp7761121.pdf

[Architect]

All of us who have been through higher education, in particular university or similar establishments, are aware of the limits of analysis and likewise the length of study required to even begin to understand issues like structural engineers. I therefore often wonder about just how far through the educational system many of the troothers have been....

...of course that doesn't explain AE911Truth, but every profession has a few imbeciles. And if they don't like that comment, they should feel free to complain to the RIBA where I will happily show them for the misguided fools that they are.

[Apollo20]

Gumboot:

You claim that:

"Likewise with the footage of UA175 hitting WTC2, I have pointed out that the margin of error is so large that calculating the deceleration of the aircraft is impossible."

If you really believe this I think you better write to Yukihiro Omika at the Kajima Corporation in Tokyo Japan and ask him how he accompished what you say is an "impossible"feat; namely the "impossible" calculation of decelerations using a video of Flight 175 as described in his 2005 paper entitled "Structural Response of World Trade Center under Aircraft Attacks"

Then, while you are at it, I think a letter to the ASCE Editors of the Journal of Structural Mechanics would be in order asking how they could be so inept as to accept a paper for publication that includes an impossible calculation!

[deep]

We see it often in regards to NIST's study of the WTC collapses. They have stated repeatedly that it is simply not possible to model the actual collapse progression, as the number of variables is too high.


I see no reason to doubt that; however, who's to say that a model of that complexity was the right approach in the first place? If they decided to model everything at the molecular level, and then announced that it was too complex to finish, would that also be acceptable?

To put it simply, choosing the wrong tool for a job is not a valid excuse for failing to do the job at all. You don't need to be an architect or an engineer to understand that.

[rwguinn]

I see no reason to doubt that; however, who's to say that a model of that complexity was the right approach in the first place? If they decided to model everything at the molecular level, and then announced that it was too complex to finish, would that also be acceptable?

To put it simply, choosing the wrong tool for a job is not a valid excuse for failing to do the job at all. You don't need to be an architect or an engineer to understand that.

Ok, mr know-itt-all--tell those of us who do this (Rule10) for a living just WTF tool we should use?
Put up orshut up. I'm tired of this (Rule10)!

Since what I do is modeling, using numerous tools--IDEAS, NASTRAN, PATRAN, ANSYS-- so I am thus unqualified to even think about modeling a building--I am familiar with how the tools work. I even know much of the math behind it, which, again, disqualifies me from commenting.
Every node in a model has 6 degrees of freedom.
BEAMS:In order to determine bending, a beam mustbe divided into 2 beams, which have at least 3 nodes. If you are doing non-linear analysis, oh, say--post failure stuff--you need 3 nodes per beam.
Now, how many beams do you need in the building--remembering that each member on a truss must have at least 5 nodes.

PLATES. Each plate must have a node where it ties into a beam. Between beams, to define the structure, you must have at least one more plate (shell) element. Each shell must have at least 8 nodes, as many as 5 of which may be shared with beam nodes.

Now add up your beams and multiply by 5. Add up your shells and multiply by 8. Now you have the total degrees of freedom=ndof.
You are going to have 2 arrays (Matrices) that are ndof X ndof in size. I would surmise that you could model 1 floor of the building with a mere million (1.0e6) degres of freedom--but it won't be very accurate.
Now you have to decompose them, invert the stiffness--and get a deflection. Now, compare that deflection to the deflection that youn had before, adjust the stiffness based on the stress/strain curve for each material involved, and repeat until the answer converges.
Now you can go to the next time step. Repeat.
At some point, the flatness of the stress/strain curve and the large deflection that a failed element allows makes the programs mathematically choke.
It doesn't matter what program--they all use the same math. Once parts rupture, the load paths have to be re calculated, and the program chokes--it will not converge.
And, amazingly enough, the further into failure you go, the more dof you need. and the mode dof, the finer the model. and the finer the model, the more time it takes--and the more failures you get.

[deep]

Ok, mr know-itt-all--tell those of us who do this (Rule10) for a living just WTF tool we should use?
Put up orshut up. I'm tired of this (Rule10)!


With any model (even the type you described), you're going to lose some detail, unless you're modeling everything at the molecular level. You must choose an abstraction point, and work from there. NIST chose the wrong abstraction point, because they could only get so far into the investigation before throwing their hands up and saying, "we can't continue". They knew beforehand what they were tasked with, and they chose the abstraction point with that in mind -- that's an unacceptable display of incompetence.

That would be like me saying, "I'm trying to figure out if I have enough gas in my car to get to Springfield, but since I don't know how many red lights I'm going to run into, there's no way for me to know". I might not be able to figure it out at that level of precision, but I can get a pretty good idea by ignoring details like which lane I'm in, how many red lights I hit, etc.

[Architect]

But its academic, because those of us who deal with structures on a professional basis know that progressive collapse would be expected following the initial failure. Where are you going with this?

[rwguinn]

But its academic, because those of us who deal with structures on a professional basis know that progressive collapse would be expected following the initial failure. Where are you going with this?I guess he thinks (like one ijit I work with) that rigid body dynamics can explain the behavior of flexible body strain.

[technoextreme]

Gumboot:

You claim that:

"Likewise with the footage of UA175 hitting WTC2, I have pointed out that the margin of error is so large that calculating the deceleration of the aircraft is impossible."

If you really believe this I think you better write to Yukihiro Omika at the Kajima Corporation in Tokyo Japan and ask him how he accompished what you say is an "impossible"feat; namely the "impossible" calculation of decelerations using a video of Flight 175 as described in his 2005 paper entitled "Structural Response of World Trade Center under Aircraft Attacks"


Ehhh... He uses the word approximatly in that paper. It renders your whole entire argument moot.

[gumboot]

Gumboot:

You claim that:

"Likewise with the footage of UA175 hitting WTC2, I have pointed out that the margin of error is so large that calculating the deceleration of the aircraft is impossible."

If you really believe this I think you better write to Yukihiro Omika at the Kajima Corporation in Tokyo Japan and ask him how he accompished what you say is an "impossible"feat; namely the "impossible" calculation of decelerations using a video of Flight 175 as described in his 2005 paper entitled "Structural Response of World Trade Center under Aircraft Attacks"

Then, while you are at it, I think a letter to the ASCE Editors of the Journal of Structural Mechanics would be in order asking how they could be so inept as to accept a paper for publication that includes an impossible calculation!



Do you have contact information for these people?

I'm not a physicist, or a scientist, but I studied cinematography and I'm well aware of the problems with calculating the deceleration of UA175 using the footage we have.

I'd be interested to see what their calculations are, how they come up with them, and if they take into account the enormous margin of error.

You're arrogance is noted once more, Apollo20, however this is one rare topic where my knowledge far surpasses yours. Don't let your arrogance get ahead of you.

-Gumboot

[rwguinn]

Ehhh... He uses the word approximatly in that paper. It renders your whole entire argument moot.
Assuming, as Apollo20 said, there are 6 frames from impact to the 0.2 seconds till the tail disappears.
each frame the represents then, .04 seconds. At 500kt, the aircraft is moving about 241 m/sec at impact. in .04 seconds, then, it will move 9.6m. If your resolution is 1m (a rather wild assumption), then your error potential is +/-10.4%. Fairly substantial, I'm afraid.
This small an error assumes that the aircraft travel at impact was orthogonal to the lens line of sight, and that the side of the building was exactly parallel the line of sight, in the exact center of the frame--another wild assumption, since I can see windows in the pics.

[gumboot]

For a high enough quality photo there are a couple of ways that might be posible.


You could possibly do it if the photograph was taken with celluloid and you could take a look at the celluloid cross section using a high powered microscope, but I can't see any other way it would be possible. A black and white capture device only captures luma and not chroma.

-Gumboot

[gumboot]

Assuming, as Apollo20 said, there are 6 frames from impact to the 0.2 seconds till the tail disappears.
each frame the represents then, .04 seconds. At 500kt, the aircraft is moving about 241 m/sec at impact. in .04 seconds, then, it will move 9.6m. If your resolution is 1m (a rather wild assumption), then your error potential is +/-10.4%. Fairly substantial, I'm afraid.
This small an error assumes that the aircraft travel at impact was orthogonal to the lens line of sight, and that the side of the building was exactly parallel the line of sight, in the exact center of the frame--another wild assumption, since I can see windows in the pics.



The problem is you have a pretty big margin of error before you even get an original digital image. Some factors that produce a margin of error are:

1) pixel bleed across the CCD (especially in bright conditions)
2) contrast clipping (DV has poor contrast ratios meaning a loss of detail in high contrast images)
3) Shutter speed (lower shutter speed produces motion blur)
4) Source compression (see below)
5) Field rate (affects how many discreet images of the object there is, DV is interlaced which immediately introduces a margin of error)
6) Perspective (objects further from the lens are smaller)
7) Focal Length (affects the significance of 6)

That's all just in getting the image through the lens and either onto tape or out to the transmitter.

And bear in mind the process of source compression for DV is itself a good place for picking up errors because it involves:

a) low pass filtering (passing light through a blurred lens to prevent aliasing)
b) demosaicing of raw input into RGB channels
c) colour values tweaked
d) white balance
e) gamma correction
f) chroma subsampling (4:1:1 for DV)
g) sharpening (to counter effect of low pass filtering)

At each of these steps errors can occur, especially at the chroma subsampling stage, which is why DV is considered a poor format for doing chroma keying (Ace Baker, take note). The blurring of the raw imagine followed by artificial sharpening is also an obvious place where significant errors can occur.

Once you get to that point the number of factors can multiply enormously depending on what is done with the footage. For example, did these scientists get hold of actual original footage, or were they copies? How many format changes, duplications, and other alterations did the source footage go through before reaching the scientists?

The problem is, in most of the footage the margin of error only need to be a few pixels and any results are meaningless. If we take Ace Baker's attempt, for example, a margin of error of only +/- 2 pixels rendered his entire analysis void.

With all of the factors at play, a margin of error of 10+ pixels is not out of the question.

-Gumboot

[Monza]

...We see it often in regards to NIST's study of the WTC collapses. They have stated repeatedly that it is simply not possible to model the actual collapse progression, as the number of variables is too high...


What is funny is that there are twoofers who then take the exact opposite extreme stance. Rather than complain about the lack of perfectly accurate models, they believe they can model a complex system such as the WTC with Coke cans or office supplies.

[rwguinn]

<<Snip technical stuff that makes engineering sound easy>>

With all of the factors at play, a margin of error of 10+ pixels is not out of the question.

-Gumboot

My experience is with film. That stuff Kodak turned out by the running mile. All you had to do was get the right film speed, lights, and camera position, and you didn't have any of that bleed crap...:D

My point was that if everything were perfect, a 1 meter error is position is a 10% error in velocity. The assumption that you can be accurate to within 1 meter was Dr. Greenings, not mine!

But then, He's a PhD, and I am a mere BSME, PE

[gumboot]

My experience is with film. That stuff Kodak turned out by the running mile. All you had to do was get the right film speed, lights, and camera position, and you didn't have any of that bleed crap...:D

:D

Kodak are currently running a marketing campaign for the 35mm motion film stock.

"On film you're already in high-def - with pixels to spare!"



My point was that if everything were perfect, a 1 meter error is position is a 10% error in velocity. The assumption that you can be accurate to within 1 meter was Dr. Greenings, not mine!


Absolutely, you're quite right. When I saw Ace Baker counting pixels I was bemused. If Dr Greening is right, and serious scientists are doing the same, I'm absolutely stunned.

Counting pixels in DV footage to determine an object's acceleration is the photographic interpretation equivalent of building model WTC towers out of chicken wire. It displays a total lack of understanding of the field.

(It's important here to point out that trying to calculate an object's acceleration in a frame is a whole magnitude less accurate than calculating an object's average velocity in a frame)

-Gumboot

[beachnut]

With any model (even the type you described), you're going to lose some detail, unless you're modeling everything at the molecular level. You must choose an abstraction point, and work from there. NIST chose the wrong abstraction point, because they could only get so far into the investigation before throwing their hands up and saying, "we can't continue". They knew beforehand what they were tasked with, and they chose the abstraction point with that in mind -- that's an unacceptable display of incompetence.

That would be like me saying, "I'm trying to figure out if I have enough gas in my car to get to Springfield, but since I don't know how many red lights I'm going to run into, there's no way for me to know". I might not be able to figure it out at that level of precision, but I can get a pretty good idea by ignoring details like which lane I'm in, how many red lights I hit, etc.
You kind of covered both side of this argument very poorly, and I was just scanning your stuff, not really reading for fear I may explode. I think I may be correct.

[beachnut]

:D

Kodak are currently running a marketing campaign for the 35mm motion film stock.

"On film you're already in high-def - with pixels to spare!"

Absolutely, you're quite right. When I saw Ace Baker counting pixels I was bemused. If Dr Greening is right, and serious scientists are doing the same, I'm absolutely stunned.

Counting pixels in DV footage to determine an object's acceleration is the photographic interpretation equivalent of building model WTC towers out of chicken wire. It displays a total lack of understanding of the field.

(It's important here to point out that trying to calculate an object's acceleration in a frame is a whole magnitude less accurate than calculating an object's average velocity in a frame)

-Gumboot
You are right, you can not get the deceleration. They can only get an estimate at best.

[geni]

You could possibly do it if the photograph was taken with celluloid and you could take a look at the celluloid cross section using a high powered microscope, but I can't see any other way it would be possible. A black and white capture device only captures luma and not chroma.

-Gumboot

Dyes of different colours don't normaly have exactly the same brightness (this is an important consideration when testing to see if animals can see colour).

[gumboot]

Dyes of different colours don't normaly have exactly the same brightness (this is an important consideration when testing to see if animals can see colour).


Both film and (especially) digital crush the luma to fit within the black-white limits of the contrast curve. The exposure level is then altered to get a "best fit" depending on the brightness of the subject.

As such, there's very little direct link between the brightness of a given hue and the luma of that particular item within the image. With a sufficient change in exposure everything in the frame can be straight white or straight black - exactly the same luma.

-Gumboot

[Hokulele]

Apollo20, are you sure on the title of that article and where it was published? I could not find any references to it on the Kajima technical document list for 2004, 2005, or 2006, and the ASCE doesn't seem to have a "Journal of Structural Mechanics".

http://www.kajima.co.jp/tech/katri/technical/annual/vol_53/index.html
http://pubs.asce.org/journals/

[fezzic]

Apollo20, are you sure on the title of that article and where it was published? I could not find any references to it on the Kajima technical document list for 2004, 2005, or 2006, and the ASCE doesn't seem to have a "Journal of Structural Mechanics".

http://www.kajima.co.jp/tech/katri/technical/annual/vol_53/index.html
http://pubs.asce.org/journals/

ASCE Books (http://cedb.asce.org/cgi/WWWdisplay.cgi?0500041) <--Try here.

[Hokulele]

ASCE Books (http://cedb.asce.org/cgi/WWWdisplay.cgi?0500041) <--Try here.


Excellent! Thanks, I will see if I can access the full text of the article when I am at work tomorrow.



*Whacks the stupid search engine on her home computer*

[technoextreme]

Excellent! Thanks, I will see if I can access the full text of the article when I am at work tomorrow.



*Whacks the stupid search engine on her home computer*
I all ready did. It's really not that interesting and it doesn't show how he got that information.

[AZCat]

Ok, mr know-itt-all--tell those of us who do this (Rule10) for a living just WTF tool we should use?
Put up orshut up. I'm tired of this (Rule10)!

Since what I do is modeling, using numerous tools--IDEAS, NASTRAN, PATRAN, ANSYS-- so I am thus unqualified to even think about modeling a building--I am familiar with how the tools work. I even know much of the math behind it, which, again, disqualifies me from commenting.
Every node in a model has 6 degrees of freedom.
BEAMS:In order to determine bending, a beam mustbe divided into 2 beams, which have at least 3 nodes. If you are doing non-linear analysis, oh, say--post failure stuff--you need 3 nodes per beam.
Now, how many beams do you need in the building--remembering that each member on a truss must have at least 5 nodes.

PLATES. Each plate must have a node where it ties into a beam. Between beams, to define the structure, you must have at least one more plate (shell) element. Each shell must have at least 8 nodes, as many as 5 of which may be shared with beam nodes.

Now add up your beams and multiply by 5. Add up your shells and multiply by 8. Now you have the total degrees of freedom=ndof.
You are going to have 2 arrays (Matrices) that are ndof X ndof in size. I would surmise that you could model 1 floor of the building with a mere million (1.0e6) degres of freedom--but it won't be very accurate.
Now you have to decompose them, invert the stiffness--and get a deflection. Now, compare that deflection to the deflection that youn had before, adjust the stiffness based on the stress/strain curve for each material involved, and repeat until the answer converges.
Now you can go to the next time step. Repeat.
At some point, the flatness of the stress/strain curve and the large deflection that a failed element allows makes the programs mathematically choke.
It doesn't matter what program--they all use the same math. Once parts rupture, the load paths have to be re calculated, and the program chokes--it will not converge.
And, amazingly enough, the further into failure you go, the more dof you need. and the mode dof, the finer the model. and the finer the model, the more time it takes--and the more failures you get.

But...but...but... We have big computers, and your matrices are sparse (I think), and it's just math, and isn't this what you engineers are paid to do anyway? Why can't you just give them some numbers?


I think that along with your excellent points, it is also important to note that just because you can model something doesn't mean you should. What would a post-collapse initiation simulation of the towers tell us? Would we be able to use the information from that to improve future building designs? Probably not. It is unlikely that such a simulation (if possible) would duplicate the actual collapses, because of (as has been pointed out several times) the quantity of variables and the nonlinearity of the system. Even if some sort of a "Monte Carlo" simulation was run to get an idea of the probabilistic response of the system, I don't think it would tell us anything useful. All the effort in the building industry is in preventing collapse initiation (or delaying it in the case of catastrophic events) so why should anyone care what happens afterwards?

[Sabrina]

An ounce of prevention is worth a pound of cure, as they say. In this case, I think they are absolutely right to focus on what CAUSED the collapse versus what happened after the collapse INITIATED. Once it initiated, nothing short of an act of God (or FSM) was going to stop it. So learning what CAUSED it helps to PREVENT it in the future. Why is it twoofers can't grasp that simple concept?

[JonnyFive]

All told, the towers did pretty well considering the damage they sustained. They were both able to stay up for roughly an hour with massive damage to their support structure and raging fires around the point of impact. It could have much, much worse if they'd collapsed after, say, twenty minutes before most of the people inside had a chance of evacuate.

[rwguinn]

But...but...but... We have big computers, and your matrices are sparse (I think), and it's just math, and isn't this what you engineers are paid to do anyway? Why can't you just give them some numbers?


I think that along with your excellent points, it is also important to note that just because you can model something doesn't mean you should. What would a post-collapse initiation simulation of the towers tell us? Would we be able to use the information from that to improve future building designs? Probably not. It is unlikely that such a simulation (if possible) would duplicate the actual collapses, because of (as has been pointed out several times) the quantity of variables and the nonlinearity of the system. Even if some sort of a "Monte Carlo" simulation was run to get an idea of the probabilistic response of the system, I don't think it would tell us anything useful. All the effort in the building industry is in preventing collapse initiation (or delaying it in the case of catastrophic events) so why should anyone care what happens afterwards?

One of the points the twoofers and their cheerleaders always gripe about is the detail. To actually model the colapse the way they want, one would have to have the same number of nodes as total individual units of material remaining in the pile and scattered around NYC.
Each large beam remaining would require at least 3 nodes, each plate at least 8, and each dust particle 1. How many nodes is that?

Additionally, yes, the original matrix is pretty sparse, but it is coupled. (i.e., row n, col n has terms at row n, col. m, and row m col n, and many more, likely).At rupture, it becomes uncoupled. so now you have to completely rebuild the arrays, and capture the dynamics of that particle, which may have more than one node, and its interaction with other particles and indeed the remaining coupled structure, ad infinitum.
How many roads must a man walk down? how high is up?

[Newtons Bit]

One of the points the twoofers and their cheerleaders always gripe about is the detail. To actually model the colapse the way they want, one would have to have the same number of nodes as total individual units of material remaining in the pile and scattered around NYC.
Each large beam remaining would require at least 3 nodes, each plate at least 8, and each dust particle 1. How many nodes is that?

Additionally, yes, the original matrix is pretty sparse, but it is coupled. (i.e., row n, col n has terms at row n, col. m, and row m col n, and many more, likely).At rupture, it becomes uncoupled. so now you have to completely rebuild the arrays, and capture the dynamics of that particle, which may have more than one node, and its interaction with other particles and indeed the remaining coupled structure, ad infinitum.
How many roads must a man walk down? how high is up?

I don't even want to think about how complicated that is. Just doing a simple 6 d.o.f. beam anyalsis by hand is enough to make my brain want to escape from my cranium and go on vacation to France.

[Newtons Bit]

An ounce of prevention is worth a pound of cure, as they say. In this case, I think they are absolutely right to focus on what CAUSED the collapse versus what happened after the collapse INITIATED. Once it initiated, nothing short of an act of God (or FSM) was going to stop it. So learning what CAUSED it helps to PREVENT it in the future. Why is it twoofers can't grasp that simple concept?

Well, the truthers can't get a grasp on the fact that there is absoluetly no way for the collapse to be stop once started. And unfortunately, people like Gordon Ross have done wonders to perpetuate that myth. So long as the truthers believe that the towers should have self-arrested, they'll never except anything by NIST.

Of course, if they do accept that the towers would have collapsed, they probably will find new reasons to reject all studies into why it started.

[Architect]

Ach man, whit de ye ken? Yer only a qualified structural engineer!

[AZCat]

One of the points the twoofers and their cheerleaders always gripe about is the detail. To actually model the colapse the way they want, one would have to have the same number of nodes as total individual units of material remaining in the pile and scattered around NYC.
Each large beam remaining would require at least 3 nodes, each plate at least 8, and each dust particle 1. How many nodes is that?

Somewhere this side of infinity. Don't forget that there are interactions between these various particles (like gravity), so you'll need to model that if you want complete accuracy. I'm not sure if anyone has ever tried to model even for an ideal gas the particle-to-particle interactions.

Additionally, yes, the original matrix is pretty sparse, but it is coupled. (i.e., row n, col n has terms at row n, col. m, and row m col n, and many more, likely).At rupture, it becomes uncoupled. so now you have to completely rebuild the arrays, and capture the dynamics of that particle, which may have more than one node, and its interaction with other particles and indeed the remaining coupled structure, ad infinitum.

Math, math, math - this is all excuses! The NIST had $20 million! That's enough to model every particle twice!!11! [/CTer]

How many roads must a man walk down? how high is up?

42.