NIST Computer Models?

[Panoply_Prefect]

Hi!

Several times I've come across the claim that NIST:

1. Doesn't release its computer models or

2. Doesn't release its computer simulations.

I was wondering if you could help me out with two things:

a) Has NIST given any explanation for not releasing this (simulation or model whichever applies)?

b) Be the devils advocat (if that proverb exists in english...) and tell me why its relevant, e.g, what part of the computer simulations/models didn't NIST release, and what does it prevent independent (as not from NIST) researchers/truthers/whatever finding out/verifying about the NIST conclusions. The truthers I've come across claims this makes the results of the NIST report non-reproducable (yes I know, I should direct that question to the truthers, but fact is most who claim this hasn't really read it. Still I find that I do most of my arguing for the fencesitters).

Cheers,
SLOB

[qarnos]

There was a big discussion about this a while ago (6 months or so). From what I remember, the NIST models aren't exactly something you could run on your PC.

[rwguinn]

Hi!

Several times I've come across the claim that NIST:

1. Doesn't release its computer models or

2. Doesn't release its computer simulations.

I was wondering if you could help me out with two things:

a) Has NIST given any explanation for not releasing this (simulation or model whichever applies)?

b) Be the devils advocat (if that proverb exists in english...) and tell me why its relevant, e.g, what part of the computer simulations/models didn't NIST release, and what does it prevent independent (as not from NIST) researchers/truthers/whatever finding out/verifying about the NIST conlusions. The truthers I've come across claims this makes the results of the NIST report non-reproducable (yes I know, I should direct that question to the truthers, but fact is most who claim this hasn't really read it. Still I find that I do most of my arguing for the fencesitters).

Cheers,
SLOB

Drop 'em an Email. Unless they are propriatary, they should be releasable.
IIRC, there is not "a" model, but several, to cover various conditions.
I know that I would not have a single model to cover impact, fire propigation, and the quasi-steady state fire conditions of the last half hour or so before collapse.
It would probably require at least 2 models to cover the first 20 seconds of the initial impact, not counting the fire- 1 dynamic, and another static.
ETA_ Not that they will do you a whole lot of good, unless you have access to a Cray or other "supercomputer" with lots and lots of VM and plenty of scratch space. They are big, complicated, and require specialized programs to run. Be prepared to spend upwards of $50,000 US just for the right to use the programs--which ARE propriatary, commercial, and also come under the International Trade in Arms Regulations, other export restrictions, and carry big penalties for exporting them--and which belong to companies which take Copyright infringement very seriously...

[jhunter1163]

Sure I'll request 'em. And as soon as I get a Cray, I'll run 'em and let you know how it goes.

[Panoply_Prefect]

So NIST is bascially not refusing to release something that is needed to duplicate their work, but they are using proprietary software and thats the catch?

/S

[defaultdotxbe]

So NIST is bascially not refusing to release something that is needed to duplicate their work, but they are using proprietary software and thats the catch?

/S
thats one possibility, id say email them and ask, wtc@nist.gov

[rwguinn]

So NIST is bascially not refusing to release something that is needed to duplicate their work, but they are using proprietary software and thats the catch?

/S
Not so much that they are using prop. software--anybody in the US can buy or lease the software--If you don't know what the model data means, and don't have the software, it won't do you any good.
The model is nothing but lots and lots of data--In NASTRAN they would be GRID, SPC, LOAD, CBAR, CBEAM. CQUAD, CONM, RBE2, RBE3, PSHELL, PBAR, PBEAM, etc--
The input to the modeling program (Which is what you get when you ask for "the model" ), is absolutely meaningless if you don't know the code, and even if you do, it is overwhelming. With something as large as this undoubtedly is, you probably won't be able to even plot it by hand in anything like a lifetime...

[Panoply_Prefect]

thats one possibility, id say email them and ask, wtc@nist.gov

I've tried that, but the mail bounced with a:
"
... while talking to spamav2mx.nist.gov>>> DATA
<<< 550 5.7.0 Local Policy Violation"


Damn it, it looks like Im spam.

/S

[WildCat]

I've tried that, but the mail bounced with a:
"

... while talking to spamav2mx.nist.gov>>> DATA

<<< 550 5.7.0 Local Policy Violation"


Damn it, it looks like Im spam.

/S
Maybe you shouldn't use v1agra4cheep@hotmail.com as your email address.

[chippy]

There's really no reason to redo the work when it's already been done. The major results can be found here:

http://wtc.nist.gov/reports_october05.htm

[Panoply_Prefect]

There's really no reason to redo the work when it's already been done. The major results can be found here:

http://wtc.nist.gov/reports_october05.htm

Im not even close to being enough qualified to judge what could be achieved from gaining access to NIST computer models. However since I keep running into the claim, I need to someway understand it and deal with it. Granted, most truthers who use it (the claim that is) haven't a clue what they are asking, and the simplest way to deal with it would be to just counter-ask them what they mean. But that won't make the claim go away and actually knowing the answer would IMHO make a better case.

Cheers,
SLOB

[T.A.M.]

[Lucifers promoter mode]

If we could get a hold of NISTs computer models, we could point out each and every miniscule error, making their entire report, based upon it, useless, and in so doing, promote our own agenda that 9/11 was an inside job...

[/Lucifers promoter mode]

TAM:)

[gumboot]

I suspect the CTers think the "NIST computer model" is this neat 3D animation of the buildings collapsing.

-Gumboot

[defaultdotxbe]

I suspect the CTers think the "NIST computer model" is this neat 3D animation of the buildings collapsing.

-Gumboot
i always figured they think NIST fired up a custom NWO version flight simulator 2004, crashed a plane into the WTC and waited for the building to collapse, when it didnt they wrote a "collapse patch" and tried again

[rwguinn]

i always figured they think NIST fired up a custom NWO version flight simulator 2004, crashed a plane into the WTC and waited for the building to collapse, when it didnt they wrote a "collapse patch" and tried again

From their writings, I believe that you are most probably correct.
As end-users of all the simulations out there, they are totally unaware of the effort that goes into something relatively simple, like MS Flight Simulator.
Flight sims have been around a very long time, and enven in the 1970's were truly crude by comparison with what's available today. And in those, all they have to do is set up the parameters for the aero data---stall speed vs AOA, Stall speed vs bank angle, roll rates, climb rate at power settings, and the like. The heavy-duty computing power comes in with the scenery and all.
Engineering sims are not at all like that. A fair amount of hand-calculation is involved, even after the sim is run, and the you have to figure out if the effects you calculate are real, or if they artifacts based on the finite element restrictions, or asssumptions, if your boundary conditions are reasonable...
It is a taks and a half. Animating the situation is often a luxury that we cannot affort, in terms of time, computing power, and money.
Even should they obtain the models and the processing power necessary to run them, they would have no way to determine what happened, and would therefore declare it a fake.

[T.A.M.]

I believe in the impact models there were a total of 2.5 Million nodes, of which it was split 1.5M for the tower sections (several floors only), 800,000 for the aircraft.

Thats alot of processing, even over a time period as short as 0.7seconds of simulation.

TAM:)

[beachnut]

Hi!

Several times I've come across the claim that NIST:

1. Doesn't release its computer models or

2. Doesn't release its computer simulations.

I was wondering if you could help me out with two things:

a) Has NIST given any explanation for not releasing this (simulation or model whichever applies)?

b) Be the devils advocat (if that proverb exists in english...) and tell me why its relevant, e.g, what part of the computer simulations/models didn't NIST release, and what does it prevent independent (as not from NIST) researchers/truthers/whatever finding out/verifying about the NIST conclusions. The truthers I've come across claims this makes the results of the NIST report non-reproducable (yes I know, I should direct that question to the truthers, but fact is most who claim this hasn't really read it. Still I find that I do most of my arguing for the fencesitters).

Cheers,
SLOB
Why is there a problem with the NIST model. The model was not done to prove CT idiots were wrong. Read why NIST did the models and did the work.

The truthers claim they have experts, they should do their own model to figure out they have no facts.

You do not need to model the WTC to know that it fell from impacts and fires; do you? If you do, you may be a truther.

[Panoply_Prefect]

Why is there a problem with the NIST model.

Ummm... I have no problem with the NIST model.


The model was not done to prove CT idiots were wrong. Read why NIST did the models and did the work.

Im not really questioning why NIST did the models or their work (And to add, from what I can gather they did a good job)


You do not need to model the WTC to know that it fell from impacts and fires; do you? If you do, you may be a truther.

Umm... I can only stress that I dont question any of this. Im not sure how I could clarify what my point is, but I really am just trying to answer this claim. Mostly I find that ct claims are thing as ice in august, and I was suspecting this was a strawman of sorts. Thats all. I wont press this further.

EDIT: Let me put it this way. Any person who are unfamiliar with basic CT-claims who watches the John Kerry clip (http://forums.randi.org/showthread.php?t=80225)and hears the "Silverstein"-claim from that source will think that Silverstein actually ordered WTC7 to be "pulled". Coming here however would pretty fast put that quote into context and show just how CT's argue. Im just trying to do the same with this claim.

Cheers,
SLOB

[AZCat]

Hi!

Several times I've come across the claim that NIST:

1. Doesn't release its computer models or

2. Doesn't release its computer simulations.

I was wondering if you could help me out with two things:

a) Has NIST given any explanation for not releasing this (simulation or model whichever applies)?

b) Be the devils advocat (if that proverb exists in english...) and tell me why its relevant, e.g, what part of the computer simulations/models didn't NIST release, and what does it prevent independent (as not from NIST) researchers/truthers/whatever finding out/verifying about the NIST conclusions. The truthers I've come across claims this makes the results of the NIST report non-reproducable (yes I know, I should direct that question to the truthers, but fact is most who claim this hasn't really read it. Still I find that I do most of my arguing for the fencesitters).

Cheers,
SLOB

IIRC there actually is independent validation of the NIST models going on. I've seen a few abstracts from academic journals that looked at the NIST models, but I don't remember where. Perhaps someone with access to a university-level journal search engine could look them up - I no longer can. It shouldn't be hard, given the right search parameters, to turn up a number of relevant articles.

[AZCat]

It is a taks and a half. Animating the situation is often a luxury that we cannot affort, in terms of time, computing power, and money.


I had a professor once say to our class that the only reason to produce graphical output like animation or pretty multicolor images was to impress the non-technical people.

[R.Mackey]

There have been several independent validations of NIST's own models, and at least four totally independent simulation efforts.

NIST describes in the report who the model creation and checking were contracted to, and the contract announcements are on their awards page (http://wtc.nist.gov/solicitations/). The NIST report also references how their models stack up against the MIT (http://web.mit.edu/civenv/wtc/) model (relatively simple) and Weidlinger Associates (http://enr.construction.com/news/buildings/archives/021104.asp) models (fairly sophisticated), as described in chapter 9.12 of NIST NCSTAR1-2B (http://wtc.nist.gov/pubs/NISTNCSTAR1-2BChap9-11Draft.pdf). A third study was done by Exponent Failure Analysis Associates (http://www.exponent.com//wtc.html) (formerly known as Failure Analysis) for Silverstein's insurers, and was thus actually hostile to the conclusions of the Weidlinger study. There is also the work of Purdue University (http://news.uns.purdue.edu/html4ever/2006/060911.Sozen.WTC.html), and probably several other efforts that I haven't even heard about.

Of the many models used in the different steps, the most difficult is probably the fire progression. NIST is still developing its own software for fire modeling (called FDS, or Fire Dynamics Simulator), and discusses in detail how it was validated in NIST NCSTAR1-5 (http://wtc.nist.gov/pubs/NISTNCSTAR1-5Draft.pdf). Fire modeling is tough, but on the plus side, the fire progression was partially observable from the ground, unlike the details of impact and core damage.

On the basis of this rich body of research, not to mention the depth of the NIST report itself, I'm extremely comfortable with its methods and conclusions. I also have a good grasp of the uncertainties in its model -- as there will be in any model -- and can defend the position that such uncertainties do not affect the overall conclusions of the report.

I suspect the primary reason for not releasing models, aside from the fact that nobody would be able to do anything with them, is that the model information is proprietary. The element models could be reverse-engineered to produce a blueprint of the WTC buildings and a Boeing 767 aircraft. Both are the property of private companies, and the latter is probably export-controlled under ITAR, in part if not whole.

The members of the Truth Movement have not produced any reason why we should take them seriously, so I am quite happy to ignore their demands. If they were to set themselves up with a legitimate research institution, with competent researchers, funding, sign non-disclosure agreements and produce an information technology security plan, then they should be able to get access to the models. But not before.

[beachnut]

EDIT: Let me put it this way. Any person who are unfamiliar with basic CT-claims who watches the John Kerry clip (http://forums.randi.org/showthread.php?t=80225)and hears the "Silverstein"-claim from that source will think that Silverstein actually ordered WTC7 to be "pulled". Coming here however would pretty fast put that quote into context and show just how CT's argue. Im just trying to do the same with this claim.

Cheers,
SLOB
There are dolts who will take what Kerry said, which was not a darn thing about 7, and said he has no knowledge on the subject, and make it 9/11 truth stuff. They are pure dolts and will never understand a model from NIST or anyone.

The pull is a classic, it prove the brain power of the standard pull truther to be nil. Silverstein statement is a red flag only to find idiots. When someone says pull is a smoking gun, you know it means they are nuts. As now with the Kerry statement, anyone with a brain knows he said nothing to support 9/11 truth.

Good luck. You want information to fight the idiots who can not comprehend simple conversations for meaning. A few experts on the 9/11 truth side could use math and physics to confirm how the WTC failed. I have seen no experts yet on the 9/11 side explain why they may be wrong. But all the papers confirming what we saw on 9/11 explain errors and assumptions.

Good luck on your quest, but it has been five years, we can only save the uniformed from becoming involved in a mind cult of lies disguised as the truth. For the one or two who base their believe in 9/11 truth due to the fact NIST will not let others see/use/etc the model, this will help them to understand there is nothing sinister to the Model being proprietary. The truth movement manufactures lies, if you can get that across you may save a few.

[Cuddles]

One thing about models is that, aside from rwguinn points out about understanding the code, even if you do understand it, you need the software to run it. For example, some of the simulations I run use MAFIA. A .com file for a simulation can be understood fairly easily and you could at least work out what the structure and initial conditions are just from reading it. However, to actually run the simulation you need the software and a license for it which will cost upwards of £30,000 per year. Even if NIST were happy to give you absolutely all the data they have, there is no point in them making it generally available because virtually nobody in the whole world would be able to use any of it.

[lozenge124]

Hi,

I wrote the NIST about these computer models and the data sets, about a month later I received a reply directing me to submit a FOIA request:

Dear Sir,

Your request will need to be sent to the NIST Freedom of Information
Act
(FOIA) Office for processing. Instructions for submitting FOIA
requests
are available on the NIST web site at
nist.gov/admin/foia/foia.htm.

Sincerely,

WTC Investigation Team

Searching around the internet, I was able to find one person who had taken the process further and submitted the FOIA request:

kolumbus.fi/totuus/doc/foia-nist.html

he quotes from the reply:
NIST used four different software packages to run its computer models." Three of these are commercially available, and one is available for free from NIST:

* LS-DYNA "was used to conduct the analysis of the aircraft impact into the towers."
* NIST used ANSYS "to conduct the thermal analysis (taking data from FDS and applying it to the structural elements to determine how they heated over time) and to conduct the structural analysis up to the point of global instability. NIST also write[sic] some code within ANSYS that allowed the transfer of the thermal data to the structural model since these models used different finite elements."
* SAP2000 "was used to construct a model of a tower and conduct baseline structural analyses."
* "The fourth computer model runs on FDS (Fire Dynamics Simulator) and SmokeView. These are software programs developed by NIST" and are publicly available.

He apparently was also given a price of US$2,534.33 for the SAP2000 model.


Finally, a couple of points:
1) I don't think it is accurate to say that the models would be useless to most people because they do not have the computer power to run the simulations. My understanding (I am not an FEA expert) is that with FEA models you can release datasets representing the values calculated by the simulation at given points in time so that 3rd parties can look at the results without running the simulation themselves. In fact, the running of a simulation and subsequent data analysis are usually done separately so there's no reason why the datasets could not be shared.

2) There was an interesting article in new civil engineer called "WTC investigators resist call for collapse visualisation" in October 2005, which has some interesting points on this.

nceplus.co.uk/fastsearch/ArchiveArticleAssetPT/?AID=22081

WORLD TRADE Center disaster investigators are refusing to show computer visualisations of the collapse of the Twin Towers despite calls from leading structural and fire engineers, NCE has learned.

Visualisations of collapse mechanisms are routinely used to validate the type of finite element analysis model used by the investigators.
(bold added)

University of Manchester professor of structural engineering Colin Bailey said there was a lot to be gained from visualising the structural response.

"NIST should really show the visualisations; otherwise the opportunity to correlate them back to the video evidence and identify any errors in the modelling will be lost, " he said.

A leading US structural engineer said NIST had obviously devoted enormous resources to the development of the impact and fire models. "By comparison the global structural model is not as sophisticated, " he said.

"The software used has been pushed to new limits, and there have been a lot of simplifications, extrapolations and judgement calls.

"This doesn't mean NIST has got it wrong in principle, but it does mean it would be hard to produce a definitive visualisation from the analysis so far."

[T.A.M.]

Hi,

I wrote the NIST about these computer models and the data sets, about a month later I received a reply directing me to submit a FOIA request:



Searching around the internet, I was able to find one person who had taken the process further and submitted the FOIA request:

kolumbus.fi/totuus/doc/foia-nist.html

he quotes from the reply:

He apparently was also given a price of US$2,534.33 for the SAP2000 model.


Finally, a couple of points:
1) I don't think it is accurate to say that the models would be useless to most people because they do not have the computer power to run the simulations. My understanding (I am not an FEA expert) is that with FEA models you can release datasets representing the values calculated by the simulation at given points in time so that 3rd parties can look at the results without running the simulation themselves. In fact, the running of a simulation and subsequent data analysis are usually done separately so there's no reason why the datasets could not be shared.

2) There was an interesting article in new civil engineer called "WTC investigators resist call for collapse visualisation" in October 2005, which has some interesting points on this.

nceplus.co.uk/fastsearch/ArchiveArticleAssetPT/?AID=22081


(bold added)


Welcome to the forum Lozenge. good, honest, civil questions/comments.

I would suggest that the meaning of the posting would be that to most of the armchair googlescientists in the 9/11 truth movement, the release of the models and the data, would be useless. the post said it would useless to MOST, not ALL.

Has NIST replied to this article, or made a statement as to why they would not release the models and or data sets?

TAM:)

[lozenge124]

Welcome to the forum Lozenge. good, honest, civil questions/comments.

I would suggest that the meaning of the posting would be that to most of the armchair googlescientists in the 9/11 truth movement, the release of the models and the data, would be useless. the post said it would useless to MOST, not ALL.

Has NIST replied to this article, or made a statement as to why they would not release the models and or data sets?

TAM:)

Hi, thanks. Just for some background, I sometimes post in the physorg forum, but I thought I'd register here as I had some information on the computer modeling issue.

From the nce article:
NIST told NCE this week that it did not believe there is much value in visualising quasistatic processes such as thermal response and load redistribution up to the point of global collapse initiation and has chosen not to develop such visualisations.

But it said it would 'consider' developing visualisations of its global structural collapse model, although its contract with the finite element analysis subcontractor was now terminated.
As this was in Oct.2005, I don't think anything is being done at this point.

(As I cannot post links yet, I'm afraid you'll have to cut and paste my links and add www sometimes) the article is also at 911review.com/coverup/imgs/resistcall.jpg

[T.A.M.]

Well has anyone asked them to release the data for the 2.5 Million nodes, and if they did get them, would they be of any use?

TAM:)

[realitybites]

I suspect the CTers think the "NIST computer model" is this neat 3D animation of the buildings collapsing.

-Gumboot
And even then, they wouldn't believe it unless it was set to a hip-hop track with a funky beat you could dance to.

Xris-Xross'll make ya'... JUMP! JUMP!

The daddy-mack'll make ya'... JUMP! JUMP!

[slugmancs]

IIRC there actually is independent validation of the NIST models going on. I've seen a few abstracts from academic journals that looked at the NIST models, but I don't remember where. Perhaps someone with access to a university-level journal search engine could look them up - I no longer can. It shouldn't be hard, given the right search parameters, to turn up a number of relevant articles.

This article talks about the Purdue models in relation to the NIST model...
http://firechief.com/mag/firefighting_wtc_sim_backs/index.html

Still trying to find any peer reviewed articles criticizing NIST, Jones doesn't even have his article on any of the databases I've searched so far - wanted to see if there are any critiques not printed on his website.

[Cuddles]

Of course, the biggest question is has anyone actually asked for anything? From Lozenge's post it seems that at least one person has found out what models they have, but has anyone, especially those making the claims, actually gone further and actually asked for either the models or the data from them? It seems a bit pointless to discuss reasons (valid though they are) why NIST might not give things out when we do not even know for sure that they don't.