AE911 Truth plans their own model

I'll believe it when I see it if they're seriously doing their own work on this... and I'm not sure whether to laugh of do a gigantic face palm in the event that they come out with explosives not giving the results they want.
 
...and I'm not sure whether to laugh of do a gigantic face palm in the event that they come out with explosives not giving the results they want.


Actually, that would be the most amazing result they could announce, since it's so clearly not the one they want. We'd have to actually start taking them a bit more seriously if they actually showed enough integrity to publish such a result.


But I'd still bet against that ever happening, mind you!
 
Of course their model will show the results they want. What they won't show you is the forty other models that didn't come out they way they wanted while they were figuring out how to stack the deck. They'll just keep moving the "explosives" around until it produces the desired outcome, then announce that this is what the ninja demolitionists "must" have done, because it duplicated the observed collapse. As I've said, the way these models work doesn't validate the assumptions that you programmatically embed in them, such as by what means you determine that a component has failed or what phenomenon failed it.
 
I am positive that one can find a placement of explosives that will mimic the collapse of the towers or WTC 7. The structures failed, all one need do it cause it to fail again in the same, or sufficiently similar, locations and in the same order and it will come down the same way.

What AE911T will simply produce is an enormity of begged questions; who put those explosives there, and when, and how were they installed surreptitiously, and why is there no physical evidence of their existence, and why is there no documentary recordings of explosive sounds that would be generated by these explosives??????

Of greater value would be a model that demonstrates the effect on the structure due to impacts and fires. Those ARE in evidence as having occurred. If AE911T were serious about 'truth' they would simply conduct a similarly complex, or better, modeling of the structural responses than any other organization has performed,
AND
be honest about it. If it shows again that the structure fails in similar fashion to the observed collapse, then admit it.
 
I am positive that one can find a placement of explosives that will mimic the collapse of the towers or WTC 7.

As someone facetiously pointed out, all you have to do is refer to NIST's sequence of thermally-predicted failures and arrange for the same items to fail in the same order in your model, insisting that instead they were destroyed by "explosives."

The computer method here doesn't validate the reasons why you tell it a structural member has failed. It just shows you what would happen if it did. To be sure, there are criteria for failure it does know how to compute, such as if you simply load a structure to its yield point. Those constitutive relationships are from structural mechanics and materials science. The "kernel" of these algorithms doesn't have a comprehensive representation of all known physics -- just the part that produces useful and stable constitutive relationships for its original purpose.

That's why NIST had to separately consider the effects of heat and programmatically add those effects into a CSD (computational structural dynamics) model. If you anthropomorphize the model, from its perspective girders and beams and columns just "magically" disappeared or behaved in a way outside its understanding. It responded appropriately to those "magical" occurrences, but had no understanding of why they occurred. From that same approach, you can "magically" fail a column by artificially introducing a fracture or gap in what previously had been a solid object, but the program will have no way of validating that you forced that gap according to your human belief that an explosive was placed there.

We are gradually moving toward unified models for finite-element methods in different regimes, but the math is hard, iffy, and unstable and the computational requirements expand obscenely. For example, you used to have to model the structural behavior of an airplane wing according to structural constitutive relationships, and its aerodynamics according to an entirely different set of constitutive relationships that derive from the Navier-Stokes models of fluid dynamics. The models didn't combine. You couldn't model the distribution of air pressure on a wing and have the structural model flex it according to that -- you had to programmatically transfer the effect of aerodynamics to the structural model as a "magic" load. But the resulting new shape of the wing changes the aerodynamic behavior, so the gestalt effect is missing. Without treading very much further on somewhat irrelevant (and proprietary) methods, we are reaching the point where we can now confidently combine these two regimes.

What AE911T will simply produce is an enormity of begged questions...

Of course, but a coherent end-to-end picture is not how conspiracy theories work. The Magic Oracle of the computer will "confirm" that explosives must have been there, so the "details" of who put them there, when, how, and all those other important questions get swept under the carpet. Somehow it "must" have happened -- even if they can't tell you when or how -- because they supposedly have iron-clad proof that it did happen.

Conspiracy theories almost always try to leapfrog over the intermediate line of reasoning and convince you of the conclusion. "We know the conclusion has to be true because of this one smoking gun; it doesn't matter how it happened or if we're able to determine how it happened -- it happened."
 
...snip

We are gradually moving toward unified models for finite-element methods in different regimes, but the math is hard, iffy, and unstable and the computational requirements expand obscenely. For example, you used to have to model the structural behavior of an airplane wing according to structural constitutive relationships, and its aerodynamics according to an entirely different set of constitutive relationships that derive from the Navier-Stokes models of fluid dynamics. The models didn't combine. You couldn't model the distribution of air pressure on a wing and have the structural model flex it according to that -- you had to programmatically transfer the effect of aerodynamics to the structural model as a "magic" load. But the resulting new shape of the wing changes the aerodynamic behavior, so the gestalt effect is missing. Without treading very much further on somewhat irrelevant (and proprietary) methods, we are reaching the point where we can now confidently combine these two regimes.



...
I am impressed. nonlinear analysis^3.
Of course, what took 3 days in 1981 now takes 7 seconds, but you've go a LOT of CPU tied up there.
But I remember when a 500 dof model was BIG!
 
I am impressed. nonlinear analysis^3.

It's the mathematical equivalent of drinking a whole bottle of absinthe and then hitting yourself repeatedly with the empty bottle. Some portions of it can get linearized through the magic of eigenvectors, but it's still high-performance computing no matter what you need.

Of course, what took 3 days in 1981 now takes 7 seconds, but you've go a LOT of CPU tied up there.

Quite. The best computers we use for this type of engineering analysis are about the size of a tennis court. Compare this to the 1960s when the fastest computers we had were also the size of a tennis court. We have had models run for something like 15 days on those hero systems, at a cost of $2,000 a day just in electricity costs. This is uncommon, though. A bread-and-butter high-performance computer is three standard racks, 64 CPUs, eight or so cores per CPU.

These days we just chop up the problem into finer elements. It takes just as long to run the analysis as it ever did, but it's better analysis -- more faithful, more stable. There are now non-uniform methods where you can adaptively subdivide and recombine the elements. But the best implementations of that are still proprietary.

But I remember when a 500 dof model was BIG!

Indeed it was. During the Apollo program, a 20-element model for thermodynamics and heat transfer was done on the lunar module. Back then that was leading edge analysis. The most d.o.f. I've set up is on the order of 1015, but that was a proof of concept.
 
Just thinking about this point in detail, it seems to me that (a) yes, in principle it's possible, (b) it's a fundamentally flawed approach, and (c) the fundamental flaw lies in the clause "more readily replicates the observed destruction". Let's look at those one by one.
I have a mixed reaction to your post. I disagree and completely agree, all at once. :p

For example, I disagree with (a). I don't think it is in principle possible. The reason is ...

Secondly, it's fundamentally flawed because it's a false dilemma fallacy. All it can ever prove, even if it succeeds, is that one carefully constructed explosive demolition scenario is more likely than one specific fire initiated scenario. It doesn't address the possibility that the actual collapse was due to a slightly different fire initiated scenario.

... which is exactly right. It's like trying to prove a negative. One can run a gazzilion cases and someone can right say, "But what about the gazzilion +1 scenario?"

Gage, et. al. are on a fool's technical errand and a finely honed grifter errand.
 
I have a leaked memo from Richard Gage that reveals some of AE911 Truth's upcoming research topics:

1. Why All Explosive Sounds Can Only Be Created by Explosives


2. Ninja Invisibility Suits: Friend of Demo Crews


3. Aeronautical 3-Card Monty: How Flights 77 and 93 Vanished Into Thin Air


4. The FDNY’s Secret: Why “They Made That Decision” To Pull WTC 7


5. Because I Say So: The Art of Claiming There’s Evidence Without Presenting Any


6. The Zen of Footprints: How Buildings Can Fall Within Their Footprints While Simultaneously Damaging Buildings Outside Their Footprints


7. First Time In History: Why Nothing Can Happen If It’s Never Happened Before


8. Why Asbestos Insulation That Was Never Installed In WTC 2 or 7 Was Nevertheless A Reason To Destroy Them

9. What the Hell is the Definition of A Pyroclastic Cloud and How the Hell Can They Be Created By Explosives and Thermite?


10. The Value of Forgetting: How To Erase Richard Gage’s Box Experiment From Everyone’s Memory

:big: Brilliant! And nominated for the Language Award.
 
I see that a number of posts started with one by gerrycan got dumped to AAH. I am a bit disappointed about how that went. I wish posters had addressed whatever he has to say with a view to the planned AE911T FEA and what we expect to happen there.

And if someone had asked gerrycan whether he is in any way involved in that project.
 
I see that a number of posts started with one by gerrycan got dumped to AAH. I am a bit disappointed about how that went. I wish posters had addressed whatever he has to say with a view to the planned AE911T FEA and what we expect to happen there.

And if someone had asked gerrycan whether he is in any way involved in that project.
gerry didn't actually say anything that would pertain to this thread.

I seriously doubt he's involved in the model charade.
 
gerry didn't actually say anything that would pertain to this thread.

I seriously doubt he's involved in the model charade.

But he's done FEA modelling, and perhaps has some ideas on how he thinks AE911Truth will or ought to tackle this. So if y'all had not run with the derail part, and instead tried to engage him on-topic...
 
But he's done FEA modelling, and perhaps has some ideas on how he thinks AE911Truth will or ought to tackle this. So if y'all had not run with the derail part, and instead tried to engage him on-topic...
The problem is he doesn't understand the method, only the tool.

Anyone can plug meaningless numbers into a model, It takes knowledge to make the model meaningful.

He has not shown a real understanding of the problem that needs to be modeled.
 
The problem is he doesn't understand the method, only the tool.

Anyone can plug meaningless numbers into a model, It takes knowledge to make the model meaningful.

He has not shown a real understanding of the problem that needs to be modeled.
Therein lies the problem. JU mentioned it upthread. It the GIGO concept. While AE911T complains that NIST fudged data and won't divulge their inputs to the fea, I get the feeling they will either do the same thing.

I mentioned in one of the moved posts that AE911T will likely only fea the girder walk off but that NIST focused first on which column failed initially and that AE911T needs to first confirm or dispute that it was col79.

Then NIST determined that without the south side damage that a col79 failure would still progress to global failure. That is more important than girder walk off as it demonstrates a design weakness. Individual floor failure in unfought fire is common. Loss of one column progressing to global collapse less so.
 
Therein lies the problem. JU mentioned it upthread. It the GIGO concept. While AE911T complains that NIST fudged data and won't divulge their inputs to the fea, I get the feeling they will either do the same thing.

I mentioned in one of the moved posts that AE911T will likely only fea the girder walk off but that NIST focused first on which column failed initially and that AE911T needs to first confirm or dispute that it was col79.

Then NIST determined that without the south side damage that a col79 failure would still progress to global failure. That is more important than girder walk off as it demonstrates a design weakness. Individual floor failure in unfought fire is common. Loss of one column progressing to global collapse less so.

It's more than reasonable to assume the extensive fires were warping the frame and things were moving about if still connected. It's academic what might have happened in "isolation" at a region surrounding a single column... but clearly this is ignoring what was going on, and had gone on in the building. There were credible FDNY reports of warping earlier in the afternoon. You don't see any one factoring this into their FEA including NIST.

The situation was too dynamic, complex and 4D and isolation one connection is absurd really.
 
Yes, which is why even if NIST got girder movement to within 1" of complete walk off then its a close to given that a combination of other factors added up to seat/girder connection failure.
 
It might be psikeyhacker's nonsense model rebadged perhaps?

I know he hates Gage with a passion, maybe they've kissed and made up.
 
Any idea where to find the technical briefs for the last two months?

Can't find anything.

I did see this at their archived org.

AEorg20150301_zpscfvgt90h.jpg


Sometimes you just have to love YouTube.

:D :D :D
 

ISF - Join now!

Every member here is approved by hand. No bots, no spam, just people who care about evidence and honest debate.

Membership is free!

Create your free account

Back
Top Bottom